probing adsorption of divalent cations on kaolinite …...probing adsorption of divalent cations on...
TRANSCRIPT
Probing Adsorption of Divalent Cations on Kaolinite Clay Surfaces by Atomic Force
Microscopy
by
Jing Chang
A thesis submitted in partial fulfillment of the requirements for the degree of
Doctor of Philosophy
in
Chemical Engineering
Department of Chemical and Materials Engineering
University of Alberta
© Jing Chang, 2019
ii
Abstract
Kaolinite is one of the most commonly-used clay minerals in the production of paper,
plastics, and ceramics, etc. However, kaolinite is also a hard-to-remove mineral in the
processing of valuable minerals or energy resources, such as bauxite, metal sulphide ores,
and oil sands. The main problem stems from the stability of the micron-to-nanosized
kaolinite particle suspension, which is dominated by the colloidal interactions between the
charged kaolinite particles. To provide clear insights into the colloidal behaviour of
kaolinite that governs its vast applications in lab research and industry, the surface
charging properties of kaolinite particles and interactions of kaolinite with divalent cations
were studied using atomic force microscope (AFM).
The two-layer kaolinite (Al2[Si2O5](OH)4) has three kinds of external surfaces: a siloxane
(SiOSi) basal plane, an aluminium oxy-hydroxyl basal plane, and edge surfaces. To obtain
different kaolinite basal planes, kaolinite suspensions with particle sizes of 1 ~ 2 μm were
prepared and deposited on clean glass and sapphire substrates using a fast heating process.
To expose the edge surfaces of kaolinite particles, ultramicrotome cutting technique was
applied to cut a resin block with kaolinite particles sandwiched in the middle.
The interaction forces between AFM tips and different kaolinite surfaces were measured
in 10 mM KCl solutions with varying concentrations of Ca2+ and Mg2+ at different pHs.
The measured interaction forces were fitted with the classical DLVO theory to derive the
Stern potentials of various types of kaolinite surfaces. Distinct responses in the measured
Stern potentials of the two basal planes to divalent cations at pH 8 and 11 implied different
binding mechanisms of the studied divalent cations at clay-solution interfaces, which were
further explored by atomic resolution imaging. The change of the surface patterns of
iii
kaolinite basal planes was observed due to the specific adsorption of divalent cations at
high pH.
The three different surfaces of kaolinite were shown to form six possible inter-particle
associations in a kaolinite suspension. The interaction energies for these six kinds of
associations were calculated using the Stern potentials determined from the AFM force
measurements. Based on the calculation, the rheological behaviour and the settling test
results of kaolinite suspension were scientifically interpreted.
The findings in this dissertation provide fundamental understanding on the adsorption of
divalent cations at the clay-solution interfaces. The techniques used in this study might
open opportunities for more extensive explorations of features on material surfaces at
atomic level.
iv
Preface
The contents in this thesis include drafts for three manuscript that have been written and
are under proofreading by co-authors. Here, we introduce the tittles and authors of the three
papers, and their contributions.
1. Jing Chang, Bo Liu, Huaizhi Shao, Rogerio Manica, Qingxia Liu, Zhenghe Xu. Imaging
ionic distribution at kaolinite-electrolyte interface: direct evidence of specific adsorption.
Jing Chang wrote this paper, performed all the experiments and all the data analysis. Bo
Liu provided the idea of using atomic resolution imaging to probe ion adsorption. Huaizhi
Shao helped with preparing the samples. Rogerio Manica did the proofreading. Qingxia
Liu provided valuable suggestions and revised the manuscript. Zhenghe Xu supervised this
work. All the authors contributed to the discussion and commented on the manuscript.
2. Jing Chang, Huaizhi Shao, Bo Liu, Rogerio Manica, Qingxia Liu, Zhenghe Xu.
Interactions between Divalent Cations and Kaolinite Basal Planes Probed by AFM Force
Measurements. Jing Chang wrote this paper, performed all the experiments and all the data
analysis. Huaizhi Shao helped with preparing samples and background electrolytes. Bo Liu
and Rogerio Manica wrote the Matlab codes. Rogerio Manica also helped with the
proofreading. Qingxia Liu provided valuable suggestions to the manuscript. Zhenghe Xu
supervised this work. All the authors contributed to the discussion and commented on the
manuscript.
3. Jing Chang, Huaizhi Shao, Bo Liu, Rogerio Manica, Qingxia Liu, Zhenghe Xu.
Understanding Rheology and Settling Behaviors of Kaolinite Suspension through Direct
v
Interaction Force Measurement Using AFM. Jing Chang wrote this paper, performed all
the experiments and all the data analysis. Huaizhi Shao helped with preparing samples and
background electrolytes. Bo Liu and Rogerio Manica wrote the Matlab codes. Qingxia Liu
provided valuable suggestions to the manuscript. Zhenghe Xu supervised this work. All
the authors contributed to the discussion and commented on the manuscript.
vi
With great power comes great
responsibility.
-- “Spider-Man”, by Stan Lee
vii
Acknowledgments
First and foremost, I wish to give my utmost gratitude to my supervisor, Professor Zhenghe
Xu for giving me this valuable opportunity to fulfill my Ph.D study at the University of
Alberta. I would like to express my heartful appreciation to Professor Qingxia Liu for being
my co-supervisor. The intelligent guidance and continuous support from Professor
Zhenghe Xu and Professor Qingxia Liu enabled me to gain research skills, develop critical
thinking, maintain proper attitude and finally achieve my academic goal.
Special appreciation to my colleague, Bo Liu, for his friendship and giving me priceless
idea on my research project.
I thank Dr. Rogerio Manica, for his help on the Matlab programming and all his
encouragement.
Many thanks to Huaizhi Shao, my colleague, who helped me a lot on hours and hours of
experiments with his patience.
I devoted my sincere gratitude to Ms. Jie Ru and Mr. James Skwarok, our technologists,
and Dr. Chen Wang, for their training on grasping essential skills and offering me their
valuable experience. I would also like to thank Ms. Lisa Carreiro for her assistance in
managing my study here.
I thank all my colleagues from the research groups of Professor Zhenghe Xu and Professor
Qingxia Liu, for their help on providing technical support, sharing knowledge and
discussing my project.
The financial support from China Scholarship Council is acknowledged.
My accomplishment and achievements are owed to all my friends, for their company,
support and understanding. My profound thanks to my parents, Huiping Zhang and Yinkui
viii
Chang, for their unconditional love and their absolute faith in me even during the hardest
time.
ix
Table of Contents
Chapter 1 Introduction..................................................................................................... 1
1.1 Background and motivations ................................................................................... 1
1.2 Objectives and scope of the thesis........................................................................... 2
1.3 Organization of the dissertation .............................................................................. 3
Chapter 2 Literature Review ........................................................................................... 5
2.1 An overview of bitumen extraction ......................................................................... 5
2.1.1 Fundamentals of water-based bitumen extraction process ............................. 5
2.1.2 Bitumen extraction process influence factors ................................................. 6
2.1.3 Oil sands tailings .......................................................................................... 10
2.1.4 Tailings management methods ..................................................................... 11
2.2 An introduction to clay minerals ........................................................................... 13
2.2.1 Definition of clay and clay minerals ............................................................. 13
2.2.2 Applications of clay minerals ....................................................................... 15
2.2.3 The anisotropic surface property of clay minerals ....................................... 16
2.2.4 Investigation techniques for surface property of clay minerals .................... 20
2.2.5 Probing anisotropic surface property of clay minerals using AFM .............. 22
2.2.6 Influence of anisotropic surface property ..................................................... 25
2.3 Kaolinite ................................................................................................................ 27
2.3.1 Background information for kaolinite .......................................................... 27
x
2.3.2 Structure of kaolinite .................................................................................... 28
2.3.3 Particle aggregation ...................................................................................... 30
2.4 Ion distribution at a solid-liquid interface ............................................................. 31
2.4.1 Electric double layer theory .......................................................................... 32
2.4.2 Debate on ion adsorption mechanisms ......................................................... 32
2.4.3 Atomic resolution imaging ........................................................................... 33
Chapter 3 Materials and Methods................................................................................. 35
3.1 Materials ................................................................................................................ 35
3.2 Sample preparation ................................................................................................ 36
3.2.1 Kaolinite basal planes ................................................................................... 37
3.2.2 Kaolinite edge surfaces ................................................................................. 38
3.3 Atomic force microscopy ...................................................................................... 38
3.3.1 AFM force measurement .............................................................................. 38
3.3.2 AFM tip evaluation ....................................................................................... 40
3.3.3 Atomic resolution imaging by AFM ............................................................ 41
3.4 Theoretical model .................................................................................................. 42
Chapter 4 Determination of Stern Potentials of Kaolinite Surfaces: Effect of Divalent
Cations and Solution pH ................................................................................................ 48
4.1 Introduction ........................................................................................................... 48
4.2 Results and discussion ........................................................................................... 49
xi
4.2.1 AFM tip calibration ...................................................................................... 49
4.2.2 Characterization of kaolinite samples ........................................................... 59
4.2.3 Effect of divalent cations on Stern potential of kaolinite Si-basal planes .... 63
4.2.4 Effect of divalent cations on Stern potential of kaolinite Al-basal planes ... 68
4.2.5 Effect of divalent cations on Stern potential of kaolinite edge surfaces ...... 72
4.3 Summary ............................................................................................................... 74
Chapter 5 Imaging of Ion Adsorptions on Kaolinite Basal Planes ............................. 79
5.1 Introduction ........................................................................................................... 79
5.2 Results and discussion ........................................................................................... 79
5.2.1 Ion adsorption of monovalent ions ............................................................... 80
5.2.2 Ion adsorption of divalent ions ..................................................................... 82
5.3 Summary ............................................................................................................... 85
Chapter 6 Understanding Rheology and Settling Behaviors of Kaolinite Suspension
........................................................................................................................................... 86
6.1 Understanding the rheology behavior of kaolinite suspension ............................. 86
6.1.1 Calculation of interaction energy ................................................................. 86
6.1.2 Shear yield stress measurement .................................................................... 87
6.1.3 Understanding kaolinite suspension rheology behavior through interaction
energy calculation .................................................................................................. 88
6.2 Understanding the settling behavior of kaolinite suspension ................................ 92
xii
6.2.1 Settling test ................................................................................................... 92
6.2.2 Understanding kaolinite suspension settling behavior through interaction
energy calculation .................................................................................................. 93
Chapter 7 Conclusions and Future Work .................................................................... 98
7.1 Conclusions ........................................................................................................... 98
7.2 Recommendations for future work ........................................................................ 99
Bibliography .................................................................................................................. 101
Appendices ..................................................................................................................... 109
Appendix A Atomic resolution image processing .................................................... 109
Appendix B Error analysis of force fitting ................................................................ 111
Appendix C Speciation diagram ............................................................................... 113
xiii
List of Tables
Table 2.1 Tailings management methods status ............................................................... 13
Table 2.2 Distinction between clay and clay minerals24 ................................................... 14
Table 2.3 Common uses of clay minerals ......................................................................... 15
Table 2.4 Anisotropic surface properties of clay minerals studied by AFM .................... 25
Table 2.5 Summary of reported PZC and IEP for kaolinite. ............................................ 30
Table 3.1 Values used to calculate the Hamaker constants. ............................................. 44
Table 5.1 Comparison of the unit cell spacing of kaolinite basal planes .......................... 81
Table 6.1 Prediction of the pH of the maximum shear yield stress for different pairs of
surfaces from the calculated interaction energy profiles. ................................................. 91
xiv
List of Figures
Figure 2.1 Program diagram of bitumen production from oil sands by open-pit mining
technology1. ........................................................................................................................ 5
Figure 2.2 Formation of water layers in a tailings pond. .................................................. 11
Figure 2.3 Schematics of phyllosilicates with anisotropic surfaces. ................................ 17
Figure 2.4 Structure of the tetrahedral sheet27. (a) tetrahedral arrangement of Si and O, (b)
projection of tetrahedron on plane of sheet, (b) top view of tetrahedral sheet (dotted red
line: unit cell area). Large grey circles represent oxygen; small red circles, silicon. ....... 17
Figure 2.5 Structure of the octahedral sheet27. (a) octahedral arrangement of Al or Mg with
O or OH, (b) projection of octahedron in two dimensions, and (c) top view of octahedral
sheet (dotted red line: unit cell area). Large white circles represent oxygen atoms; large
grey circles represent hydroxyl groups; small blue circles, aluminum atoms. ................. 18
Figure 2.6 Relative amounts of clay minerals in oil sands deposits52. ............................. 28
Figure 2.7 Structure of kaolinite clays36. Red: oxygen; yellow: silicon; purple: aluminum;
white: hydrogen. ............................................................................................................... 29
Figure 2.8 Stairstep card-house structure of aggregated kaolinite particles53. ................. 31
Figure 3.1 X-ray diffraction (XRD) analysis of kaolinite................................................. 35
Figure 3.2 Schematics of preparing kaolinite (a) Si-basal plane, (b) Al-basal plane, and (c)
edge surfaces. .................................................................................................................... 37
Figure 3.3 Illustration of the electric double layer (shown as grey thin layer) surrounding
the AFM tip, kaolinite and substrate in electrolyte solutions with 𝜅 − 1 less than the
thickness (d) of the kaolinite particle. ............................................................................... 39
xv
Figure 3.4 SEM images of the side view of (a) a cantilever with a DNP-10 AFM tip on top
at low magnification, (b) DNP-10 AFM tip at high magnification, (c) cantilever with a
SNL-10 AFM tip on top at low magnification, and (d) SNL-10 AFM tip at high
magnification. ................................................................................................................... 41
Figure 3.5 Schematic illustration of the AFM tip-flat surface system used in the theoretical
calculation. Angles α and β are the geometrical angles for the spherical cap at the tip apex
and the conical tip body, respectively, and α + β = 90°; D refers to the separation between
the end of the tip apex and the surface of the substrate; L is the distance from the AFM tip
to the substrate; R is the radius of the spherical cap at the tip apex; r is the radius of the tip
at a given vertical position40. ............................................................................................ 42
Figure 4.1 Zeta Potential of silica wafer in 10 mM KCl solutions at pH 8 and 11 with
varying concentrations of divalent cations (Ca2+ and Mg2+, respectively) determined using
the streaming potential method. ........................................................................................ 49
Figure 4.2 Comparison of the zeta potential of Si3N4 nano particles with the Stern potential
of AFM Si3N4 tip in 10 mM KCl solutions at pH 8 and 11 containing different
concentrations of (a) Ca2+ and (b) Mg2+. .......................................................................... 51
Figure 4.3 Interaction forces between the AFM Si3N4 tip and the silica wafer in 10 mM
KCl solutions (a) at different pH, (b) at pH 8 containing different concentrations of Ca2+,
(c) at pH 8 containing different concentrations of Mg2+, (d) at pH 11 containing different
concentrations of Ca2+, and (e) at pH 11 containing different concentrations of Mg2+; and
(f) Comparison of the Stern potentials of the AFM Si3N4 tip obtained in this this study with
those reported in literature. ............................................................................................... 53
xvi
Figure 4.4 (a) Representative AFM image of a silica wafer surface; and typical interaction
forces (solid lines represent theoretical DLVO forces, open symbols represent
experimental data) between the AFM Si tip and the silica wafer in 10 mM KCl solutions
(b) at different pH, (c) at pH 8 containing different concentrations of Ca2+, (d) at pH 8
containing different concentrations of Mg2+, (e) at pH 11 containing different
concentrations of Ca2+, and (f) at pH 11 containing different concentrations of Mg2+. ... 57
Figure 4.5 SEM images of kaolinite particles deposited on (a) glass substrate and (b)
sapphire substrate after drying on a hot plate; AFM images of kaolinite particles deposited
on (c) glass substrate and (d) sapphire substrate, after the substrate was rinsed with Milli-
Q water and blown dry. ..................................................................................................... 59
Figure 4.6 Interaction forces between an AFM tip and silica wafer (a), kaolinite particles
sitting on a glass substrate (b), and particles sitting on a sapphire substrate (c) in 10 mM
KCl solutions of pH 5 (open symbols represent experimental data, and solid lines represent
theoretical DLVO forces). ................................................................................................ 61
Figure 4.7 SEM images of (a) kaolinite particles deposited on the resin surface, and (b)
kaolinite edge surfaces prepared by ultramicrotome cutting technique; and (c) AFM image
of kaolinite edge surfaces.................................................................................................. 62
Figure 4.8 Interaction forces between the AFM Si3N4 tip and the kaolinite Si-basal plane
in 10 mM KCl solutions (a) at different pH, (b) at pH 8 containing different concentrations
of Ca2+, (c) at pH 8 containing different concentrations of Mg2+, (d) at pH 11 containing
different concentrations of Ca2+, and (e) at pH 11 containing different concentrations of
Mg2+; and (f) Summary of the Stern potential of kaolinite Si-basal plane at pH 8 and 11 in
10 mM KCl solutions containing different concentrations of Ca2+ and Mg2+. ................. 64
xvii
Figure 4.9 Interaction forces between the AFM Si3N4 tip and the Al-basal plane of kaolinite
in 10 mM KCl solutions (a) at different pH, (b) at pH 8 containing different concentrations
of Ca2+, (c) at pH 8 containing different concentrations of Mg2+, (d) at pH 11 containing
different concentrations of Ca2+, and (e) at pH 11 containing different concentrations of
Mg2+; and (f) Summary of the Stern potential of kaolinite Al-basal plane at pH 8 and 11 in
10 mM KCl solutions containing different concentrations of Ca2+ and Mg2+. ................. 68
Figure 4.10 Typical interaction forces between the AFM Si tip and kaolinite edge surface
in10 mM KCl solutions (a) at different pH, (b) at pH 8 containing different concentrations
of Ca2+, (c) at pH 8 containing different concentrations of Mg2+, (d) at pH 11 containing
different concentrations of Ca2+, and (e) at pH 11 containing different concentrations of
Mg2+; and (f) summary of the Stern potentials of kaolinite edge surfaces. ...................... 72
Figure 4.11 Schematics of the adsorption of divalent cations on kaolinite surfaces at pH 8.
........................................................................................................................................... 77
Figure 4.12 Schematics of the adsorption of divalent cations on kaolinite surfaces at pH 11.
........................................................................................................................................... 77
Figure 5.1 Surface lattice structure of kaolinite: Top view of the T sheet (a), and atomic
resolution images of kaolinite Si-basal plane in Milli-Q water (b) and 10 mM KCl solution
(c) at pH 5; Top view of the O sheet (d), and atomic resolution images of kaolinite Al-basal
plane in Milli-Q water (e) and 10 mM KCl solution (f) at pH 5; with insets being high
resolution views (top) and 2D spectrum (Fast Fourier Transform) image of the same data
(bottom) (large white, solid circle represents oxygen; large, dotted circle shows the fourth
oxygen atom of each tetrahedron pointing downward; small red circle represents silicon;
large grey, solid circle represents hydroxyl; small blue circle represents aluminum). ..... 80
xviii
Figure 5.2 Atomic resolution images of kaolinite Si-basal (a), and Al-basal plane (b) in 10
mM KCl + 1 mM Ca2+ at pH 8; kaolinite Si-basal plane (c), and Al-basal plane (d) in 10
mM KCl + 1 mM Ca2+ at pH 11. ...................................................................................... 83
Figure 5.3 Schematic of AFM tip scanning surface in solution. ...................................... 84
Figure 6.1 Impact of solution pH on the shear yield stress of kaolinite suspension. ........ 88
Figure 6.2 Interaction energy/unit area between different kaolinite surfaces in 10 mM KCl
solutions at different pHs. ................................................................................................. 89
Figure 6.3 Proposed aggregate structures: (a) card house structure; (b) parallel stacking; (c)
dispersed structure. ........................................................................................................... 92
Figure 6.4 Settling performance of kaolinite suspension.................................................. 93
Figure 6.5 Impact of divalent cations on the interaction energy/unit area between different
kaolinite surfaces. ............................................................................................................. 94
Figure 6.6 Proposed aggregate structures in 10 mM KCl solutions with 0, 0.1 mM Ca2+
addition (a), with 0.5 and 1 mM Ca2+ addition (b), and with 5 mM Ca2+ addition (c). .... 96
Figure 6.7 Impact of Mg2+ on the settling performance of kaolinite suspension. ............ 97
Figure A1. Image processing procedure. (a) Original atomic resolution image after being
flattened, (b) Spectrum 2D image, (c) inverse FFT image, and (d) zoomed-in view of (c).
......................................................................................................................................... 109
Figure C1. Concentration diagram for 10-3 mol/L Ca2+ and Mg2+ . ................................ 113
xix
Nomenclature
Abbreviations
AFM atomic force microscope
BC boundary condition
CEC cation exchange capacity
EDL electrical double layer
FFT fluid fine tailings
IEP isoelectric point
ISR initial settling rate
MFT mature fine tailings
PZC point of zero charge
SEM scanning electron microscope
Symbols
A Hamaker constant, J
𝑐0 bulk concentration of the salt
𝑒 electronic charge, 1.6×10-19 C
𝐹𝐷𝐿𝑉𝑂 DLVO force, N
𝐹𝑒𝑑𝑙 electrostatic double-layer force, N
𝐹𝑣𝑑𝑤 van der Waals force, N
ℎ Planck’s constant, 6.63×10-34 J·s
𝑘𝐵 Boltzmann constant, 1.38×10-23 J/K
n refractive indexes
𝑛𝑖0 bulk concentration of electrolyte i, ions/m3
𝜈𝑒 main electronic absorption frequency in the UV region, s−1
𝑧𝑖 valence of ion i
Greek symbols
𝛾 shear rate, P
ε static dielectric constants
𝜀0 permittivity of vacuum, 8.85×10-12 C/mV
𝜂𝑃 plastic viscosity, mPa.s
κ-1 Debye length, m
σ surface charge density, C/m2
τ shear stress, Pa
𝜏𝐵 Bingham yield stress, Pa
𝜓 surface potential, V
1
Chapter 1 Introduction
1.1 Background and motivations
Oil Sands industry makes a major contribution to the economy of Canada. However, this
profitable industry has caused serious environmental problem due to the generation of oil
sands tailings, which mostly consists of kaolinite particles. The treatment of oil sands
tailings and the reclamation of land are of great significance for environmental protection.
The Alberta Energy Regulator enacted the Directive 074 to force operators to capture the
fluid fine tailings (FFT) within 1 year of deposition and fully reclaim the landscape in 5
years. Unfortunately, since this was beyond the technology at that time, enforcement of
Directive 074 was suspended in April 2015 due to difficulty in meeting all its targets.
Directive 074 was replaced by directive 085 at October, 2017. Directive 085 requires that
the fluid tailings volumes to be managed during and after mine operation, and the fluid
tailings should be progressively reclaimed during the life of a project. New directive
demands all fluid tailings that are related to a project need to be ready to reclaim within ten
years after the end of mine life of that project.
As an unsolved problem, management of oil sands tailings still draws considerable research
attention. The main problem of the tailings is that the fine particles in the tailings ponds
cannot form dense flocs and then settle down fast, which is caused by the repulsive
electrical double layer (EDL) forces among the charged fine particles. Currently, the two
most effective methods used in dewatering of oil sands tailings are flocculation and
coagulation. Flocculation uses polymers to bridge particles into large flocs, and the
polymer bridge can extend beyond the range of electrical double layer (EDL) repulsion.
2
Coagulation uses inorganic multivalent cations to depress the repulsive forces till the
attractive van der Waals forces become dominant and are able to bring and hold the
particles together.
To manipulate the settling process, the charging characteristics of the fine clay particles
should be well understood, as well as how the solution pH and the ions in the process water
affect the charging features of the particles. As the main reagent used in oil sands tailings
treatment, divalent cations and their impact on the anisotropic surface charging
characteristics of kaolinite particle became the focus of this research. The ion adsorption
mechanisms were also deeply investigated.
1.2 Objectives and scope of the thesis
The major objective of this work is to study the effect of monovalent and divalent ions on
the anisotropic surface charging properties of kaolinite. Through the study of the Stern
potential of kaolinite different surfaces under various solution conditions, possible ion
adsorption mechanisms were proposed, and the mechanisms were further elucidated by
atomic resolution imaging.
In the first part of this thesis, kaolinite samples were carefully prepared to expose the three
different kinds of surfaces of plate-like kaolinite particles. Then AFM force measurements
were performed on each kind of surface to explore the impact of background electrolyte
pH and the impact of the concentration of divalent cations at lower and higher pHs. By
analyzing the experimental data, proper ion adsorption mechanisms were proposed to
explain the effect of different cations at different solution conditions.
3
In the second part of this thesis, atomic resolution imaging was conducted to confirm the
ion adsorption mechanisms proposed in the first part. With the observation on the change
in surface lattice structure caused by adsorption of divalent cations at high solution pH,
specific adsorption was directly visualized and identified for the first time without
involving any EDL model.
In the last part of this thesis, the knowledge learnt from microscopic scale study was applied
to understand the behavior of kaolinite particles in the macroscopic processes. Using the
Stern potentials obtained in the first part of this work, the interaction energies between
different kaolinite surfaces were calculated to explain the rheology and the settling
behaviors of kaolinite suspension.
1.3 Organization of the dissertation
This thesis consists of seven chapters.
Chapter 1 provides a short statement of the background, the origin and the main objectives
of this research.
Chapter 2 presents a literature review on the most relevant background information or
previous research on anisotropic surface charging properties of kaolinite particles.
Chapter 3 gives detailed descriptions of the materials and key experimental methods used
in this study.
Chapter 4 shows the experimental data and in-depth data analysis on the Stern potential
results. Corresponding ion adsorption mechanisms were proposed according to the
experimental results.
4
Chapter 5 investigates the suggested ion adsorption mechanisms using experiments that
were designed and performed to confirm our theory.
Chapter 6 uses the fundamental knowledge we gained in the microscopic scale to
understand the kaolinite suspension behavior happens in the macroscopic level. By doing
this, we were hoping the fundamental knowledge can provide guidance for the application
in the real world.
Chapter 7 summarizes the work presented in this dissertation and gives suggestions on
future work in this research field.
5
Chapter 2 Literature Review
2.1 An overview of bitumen extraction
2.1.1 Fundamentals of water-based bitumen extraction process
Oil sands are a kind of valuable unconventional fossil fuel, which spread throughout the
world. Canada and Venezuela are the two countries that own the world’s two largest
sources of bitumen. In Canada, bitumen mostly concentrates in the province of Alberta.
Thanks to the devotion of Dr. Karl Clark on oil sands extraction, bitumen recovery can be
commercialized. The modified versions of the Clark Hot Water Extraction process (CHWE)
are still being used in the current bitumen extraction, in which bitumen is separated from
sand and clay particles by the joint efforts of mechanical energy, heat, and the presence of
surfactants and caustic addition. This process is illustrated in Figure 2.1.
Figure 2.1 Program diagram of bitumen production from oil sands by open-pit mining
technology1.
The bitumen recovery process basically obeys the following procedures2:
1) Lump size reduction. In tumblers or hydrotransport pipelines, warm water heats up the
lump surface, and the surface of lump is sheared away.
6
2) Bitumen liberation or separation. The influencing factors of the rate of this step are
process temperature, mechanical agitation, chemical additives and interfacial properties.
3) The liberated bitumen attaches to an air bubble. The occurrence of bitumen attaching to
an air bubble or engulfing an air bubble is determined by the process temperature. Under a
low temperature (< 35˚C), the attachment of bitumen to air bubble takes place, while
bitumen engulfs the air bubble in a warm (around 45 ˚C) or high (75~80 ˚C) temperature
condition.
4) Bitumen flotation. After aeration, bitumen floats up to the top of the vessel and the
bitumen froth can be recovered.
2.1.2 Bitumen extraction process influence factors
2.1.2.1 Slurry pH
In bitumen flotation, the induction time and contact angle can reflect the floatability and
flotation rate. Wang et al.3 measured the induction time in process water with different pH,
and found out that the induction time changed marginally for pH 8 and 9 and increased
significantly with a further increase in pH above 9.
The research of Liu et al.4 reported that the solution pH greatly affected both the long-range
and adhesion forces in a 1 mM KCl solution. The study showed that at low solution pH
(say, pH < 7), the repulsive force between bitumen surfaces was low and the adhesion force
was high, which was favorable for bitumen aeration but not favorable for bitumen
liberation. The research also revealed that the contact angle changed slowly at pH below 8
and decreased drastically at pH 10.
7
Besides, the generation of the surfactants from the acids present in the bitumen is pH
dependent. Increasing the slurry pH would lead to an enhanced natural surfactant
generation.
2.1.2.2 Electric Surface Potential (Zeta Potential)
The electrokinetic behavior of bitumen in aqueous media is due to the interaction of polar
groups contained by bitumen. Takamura and Chow5 used the ionizable surface group
model to explain the electrical properties of the bitumen/water interface, thus, the effect of
pH and calcium ions on bitumen displacement from sand grain could be understood.
Liu et al.6 reported the change of zeta potential of bitumen with solution pH at different
levels of calcium addition. The results showed that higher solution pH and lower calcium
concentration resulted in a system of more negative bitumen zeta potential, which is
favorable for bitumen detachment from the silica surface.
2.1.2.3 Ions
Typically, industrial recycle process water contains various types and amount of inorganic
ions, among which the dominant ions are Na+, Cl-, HCO3-, SO42- and a small portion of
divalent cations of Ca2+ and Mg2+.7 In bitumen extraction, ions in the process water have a
significant impact on bitumen recovery and bitumen froth quality.
Monovalent metal ions (Na+, K+)
In industrial recycle process water, the concentration of sodium ions can reach up to 1000
ppm8. Coagulation would not take place at a low sodium concentration (e.g., below 10
mM), while a sodium concentration higher than 50-75 mM would cause coagulation, thus
8
reducing bitumen recovery9–11. However, the affiliation of sodium ions with anions (such
as HCO3-) should also be taken into consideration.
Bicarbonate ions (HCO3-)
As a dominant anionic species in process water, bicarbonate ions can help buffer the slurry
in bitumen extraction at a mildly alkaline pH, which reduces the demand for caustic7.
Besides, the presence of bicarbonate ions precipitated Ca2+ from the extraction process
water, dispersed fine solids and reduced the bitumen-solid adhesion forces and thus
improving the quality of bitumen extraction and bitumen froth.
Zhao et al.7 found out that by adding small amount of acid (containing HCO3-) into the
water, the equilibrium of reactions (1.1) would be driven to the left side, and the happening
of reaction (1.2) could increase the concentration of CO32-, which caused the precipitation
of Ca2+. The mechanism can be seen as follows.
H2CO3 ⟺ H+ + HCO3− (1.1)
HCO3− ⟺ H+ + CO3
2− (1.2)
Ca2+ + CO32− ⟺ CaCO3 (1.3)
Divalent metal ions (Ca2+, Mg2+)
It was reported12 that, in the water-based extraction of bitumen from oil sands, Ca2+ and
Mg2+ in the process water had a great impact on the bitumen recovery. Ca2+ and Mg2+ were
found to have a negative impact on bitumen liberation by Zhao et al.13. This is due to the
increase of adhesion force and the decrease of long-range repulsive forces between silica
and bitumen by the presence of the calcium and magnesium cations.
9
In aqueous based bitumen extraction system, calcium ion (Ca2+) is one of the divalent ions
added to alter the bitumen surface properties. According to the work by Liu et al.4, calcium
ions could absorb on the bitumen surface through a chemical bond with surface carboxylic
groups (RCOO-) and thus affecting the bitumen surface properties. Besides, Ca2+ can also
compress the electrostatic double layer. Their study indicated that the addition of Ca2+
weakened the attractive hydrophobic forces and the adhesion forces between bitumen
surfaces, therefore the bitumen droplets could easily coagulate but were also easier to break
up.
Calcium ions have a negative effect on bitumen liberation and oil sands lump crumbling
due to their influence on reducing the receding rate of bitumen/liquid/solid three phase
contact line14,15. Moreover, the presence of divalent ions can inhibit the ability of NaOH
solution (pH = 11.8) to cause crumbling16.
According to the work of Xu and Masliyah17, the type of metal ions did not affect the
bitumen recovery results, it was the valence that accounted for the depression effect.
Besides, the reduction of bitumen recovery was not generated by the divalent cations alone,
but because of their presence in combination with other factors, such as the existence of
clay.
2.1.2.4 Surfactants
The addition of NaOH ionizes the organic acid in bitumen to generate surfactants, which
is proposed to be able to establish a three-phase contact point to initiate bitumen recession
from the sand grains and affect the electric surface potential of bitumen and fine solids. A
mechanical model that the surfactants can reduce the oil solution interfacial tension has
10
been proposed to explain why the released surfactants contribute to bitumen separation, yet
the model still remains to be verified8. However, it is a fact that alkaline pH environment
can improve bitumen recovery due to increased activity of surfactants, negative charges on
clay and sand surfaces, and reduced interfacial tension between bitumen droplets and
solids18.
2.1.3 Oil sands tailings
It was in 1967 when the Great Canadian Oil Sands (GCOS, now Suncor) learned that the
fine solids in tailings could not settle nor consolidate fast, so that oil sands tailings started
to become a problem because the tailings had to be stored in huge tailings ponds for a very
long time19. With the rapid expansion of oil sands industry, this problem is becoming more
and more intense, and the currently-existing tailings pounds occupy a total area of more
than 130 square kilometers20.
Oil sands tailings contain dissolved salts, organics, sands, fines (silts and clays),
unrecovered bitumen, and a large amount of water that can be recycled as process water
for the extraction process. The sand in the discharged slurry settles out quickly with the
trap of some fines and water, but some fines take decades to settle, thus forming the three
fractions as shown in Figure 2.2: (1) a clarified surface water layer (contains about 15-70
mg/L suspended solids); (2) a transition zone, in which the fines are settling; and (3) the
fluid fine tailing (FFT) zone, which has 15%-30% solids content. With time, a higher solids
content (>30%) tailings (mature fine tailing, shorted as MFT) is created with further
settling21.
11
Figure 2.2 Formation of water layers in a tailings pond.
2.1.4 Tailings management methods
Since 1967, tailings management in oil sands industry has evolved for over 50 years, which
can be divided into three overlapping development periods21:
1960s through 1980s: Waste management period. Large dams were built to store FFT, and
space was left to long-term storage.
1990s through 2000s: Creating solid tailings landscapes. R & D of this period focused on
sand capped composite tailings and water capped MFT end pit lakes.
2010s: Reducing the generation of tailings and speeding reclamation to meet new
regulatory goals.
There are five process methods22 currently being used to release water from oil sands
tailings, which are described as follows:
12
Method I: In a solid-bowl scroll centrifuge, with the addition of coagulant such as gypsum,
about 55% of the solids contents are consolidated into the centrifuge cake, which is
deposited in relatively thin lifts of 2t/m2-y in cells. After a winter freeze-thaw cycle, the
cake will reach its peak shear strength before an additional lift is placed. Alternatively, the
cake is continuously deposited into deep, in-pit deposits, the release of water can only
depend on self-weight consolidation.
Method II: In-line flocculation of FFT is used, and the flocculated slurry is discharged into
thin lifts into cells. The solids content is up to 60% after initial dewatering, and the
dewatered material can be put into the overburden cells. Evaporation and freeze-thaw will
help remove more water to form an integral part of a disposal structure.
Method III: In-line flocculated FFT is put in deep, large deposits. The water expressed from
the deposit is decanted from the surface with the aid of rim-diching of the deposit or
channels built on the surface. Further increase of the solids contents relies on the self-
weight consolidation.
Method IV: The FFT is drawn from the extraction process to be flocculated and thickened
in a mechanical thickener. The thickened tailings generated from this process are either
placed in deep deposits for further dewatering by consolidation, or discharged in thin lifts.
Method V: FFT is blended with sand slurry with high solids content, and with the addition
of flocculants or coagulants, a non-segregating mix with a moderately high sand-to-fines
ratio is formed and then discharged into a deep deposit. If the source of the fines is mature
fine tailings, the product is called composite tailings, otherwise, the product is called non-
segregating tailings if the source of the fines is thickened tailings.
13
The following Table 2.1 lists the tailings management methods that are currently in
commercial operation and that are still in R & D19.
Table 2.1 Tailings management methods status
Status Tailings management methods
In current commercial
operation
Composite tailings, thickened
tailings, in-line flocculation, thin-lift
dewatering of MFT
Ready for commercial
implementation
In-line flocculation, thin-lift
dewatering of MFT, centrifugation,
MFT spiking of whole tailings
In R& D Freeze-thaw consolidation of fine
tailings
2.2 An introduction to clay minerals
2.2.1 Definition of clay and clay minerals
Because clay scientists come from diverse disciplines, there are no uniform definitions for
clay and clay mineral so far. The latest effort in this direction was made by the joint
nomenclature committees (JNCs) of the Association Internationale pour L’Etude des
Argiles (AIPEA) and the Clay Minerals Society (CMS). ‘Clay’ was defined by JNCs as “a
naturally occurring material composed primarily of fine-grained minerals, which is
generally plastic at appropriate water contents and will harden with dried or fired”23, while
‘clay mineral’ was defined as “phyllosilicate minerals and minerals which impart plasticity
to clay and which harden upon drying or firing” 24. On top of this, Moore25 pointed out that,
although most of the minerals we have been dealing with are phyllosilicates, there are still
14
some exceptions. As researchers, first, we must separate the term ‘clay’ from ‘clay mineral’,
and the differences between clay and clay mineral are listed in Table 2.2.
Table 2.2 Distinction between clay and clay minerals24
Clay Clay mineral
Natural Natural and synthetic
Fine-grained (< 2 μm or < 4 μm) No size criterion
Phyllosilicates as principal constituents May include non-phyllosilicates
Plastic Plastic
Hardens on drying or firing Hardens on drying or firing
Phyllosilicates, which mainly includes seven groups25: the 1:1 type serpentine-kaolin; the
2:1 type talc-pyrophyllite, smectite, vermiculite, true (flexible) mica, brittle mica, and
chlorite.
Clay minerals, or more specifically, phyllosilicates, are characterized by the following
properties24: i. a layered structure with nanometer range (the layer thickness of the 1:1 (TO)
type is ~0.7 nm, and that of the 2:1 (TOT) type is ~1 nm); ii. the anisotropy of the layers
or particles; iii. the existence of different types of topographies: external basal surfaces,
edge surfaces, and internal (interlayer) surfaces; iv. the ease of surface modification (by
adsorption, ion exchange, or grafting), typically external, and often internal surfaces; v.
plasticity, and vi. hardening on drying or firing (with few exceptions).
15
2.2.2 Applications of clay minerals
Clay minerals are widely used in our daily life, modern industries, lab research, etc. The
ever-growing applications of clay minerals make them a hot topic for scientists. Table 2.3
shows the common uses of clay minerals.
Table 2.3 Common uses of clay minerals
Layer
type Group Industry field Common application Typical mineral Chemical formula
1:1 Kaolin
Paper,
plastics,
rubber
Pesticides
Pesticides
Ceramics
Filler,
Carrier, diluent
Porcelain,
Kaolinite Al2Si2O5(OH)4
2:1
Talc
Plastics,
rubber, paper
industry
Cosmetics,
pharmaceutics
Refractory
industry
Filler
Powders, pastes,
lotions
Refractories
Talc Mg3Si4O10(OH)2
Smectite
Building
industry
Pesticides
Cosmetics
Brick, roofing tile
Carrier
Bases of creams
Montmorillonite
(Na,Ca)0.33(Al,Mg)
2(Si4O10)(OH)2·nH2
O
Vermiculite
Construction
Package
Foundries
Heat insulation,
sound dissipation
Shock proof
materials, liquid
absorption
Thermal protection
Vermiculite
(Mg,Fe,Al)3(Al,Si)
4 O10(OH)24H2O
Mica
Electrical
industry
Paints
Cosmetics
Coating
Insulation
UV-, heat-stable and
under-water paints
Nacreous pigments
Corrosion proof,
polymer coating,
underseal
Muscovite
Illite
KAl2
(AlSi3O10)(OH)2
(K,H3O)(Al,Mg,Fe
)2(Si,Al)4O10[(OH)
2,(H2O)]
Chlorite Very few
industrial uses - Clinochlore
(Mg5Al)(AlSi3)O10
(OH)8
Source: adapted from Bergaya, F., 200624
However, apart from all the useful applications in industry, clay minerals, such as kaolinite,
chlorite, and smectite, can also be troublemakers in the purification of valuable mineral
16
resources and in the dewatering of tailings to achieve disposable sedimentation/
consolidation.
2.2.3 The anisotropic surface property of clay minerals
2.2.3.1 Structure of clay minerals
The structural elements of most clay minerals are based on two-dimensional arrays of Si-
O tetrahedral sheet (T) and two-dimensional arrays of Al- or Mg-O-OH octahedral sheet
(O)26. According to the ratio of T and O, phyllosilicates are classified as bilayer (1:1 type,
TO) and trilayer (2:1, TOT). In 1:1 type structure, the unit layer is made of one T sheet and
one O sheet, while the 2:1 type structure is characterized by one O sheet sandwiched
between two T sheets. Individual layers are held together through interlayer cations, van
der Waals interactions, electrostatic forces, and/or hydrogen bonding. The perfect
tetrahedron and octahedron sheets are electrically neutral, but when higher valance cations
are substituted by lower valence cations of similar sizes, (such as Si4+ is substituted by Al3+
in T, and Al3+ by Mg2+ in O), a net charge deficiency is generated. This phenomenon is
called isomorphic substitution. The charge deficiency is balanced by interstitial
compensating cations (K+, Na+, NH4+, Ca2+, Mg2+, Fe2+, etc.) to make the clay minerals
electrically neutral.
Due to the layered structure, a clay mineral consists of two different surfaces: the basal
surface and the edge surface, as shown in the Figure 2.3.
17
Figure 2.3 Schematics of phyllosilicates with anisotropic surfaces.
The tetrahedral sheet (as shown in Figure 2.4) is composed of silicon-oxygen tetrahedrons
with oxygen atoms located on the four corners of each tetrahedron while the silicon atom
sits in the center. Three out of the four oxygen atoms of a tetrahedron are shared by three
neighboring tetrahedron, and the fourth oxygen atom (the apical oxygen) pointed
downward and forms part of the adjacent octahedral sheet26.
Figure 2.4 Structure of the tetrahedral sheet27. (a) tetrahedral arrangement of Si and O, (b)
projection of tetrahedron on plane of sheet, (b) top view of tetrahedral sheet (dotted red
line: unit cell area). Large grey circles represent oxygen; small red circles, silicon.
The octahedral sheet as shown in Figure 2.5 is linked laterally by sharing octahedral edges,
in which, Al or Mg atoms are coordinated with six O atoms or OH groups which are located
around the Al or Mg atom with their centers on the six corners of an octahedron. The O
18
atoms and the OH groups lie in two parallel planes with Al or Mg atoms between these two
planes26.
Figure 2.5 Structure of the octahedral sheet27. (a) octahedral arrangement of Al or Mg with
O or OH, (b) projection of octahedron in two dimensions, and (c) top view of octahedral
sheet (dotted red line: unit cell area). Large white circles represent oxygen atoms; large
grey circles represent hydroxyl groups; small blue circles, aluminum atoms.
2.2.3.2 Layer/ surface charging mechanisms
For a phyllosilicate, the tetrahedral sheet and the octahedral sheet join together to form a
layer, which can be electrically neutral or negatively charged. A electrically neutral layer
case occurs when (i) two octahedral sites are occupied by trivalent cations (usually Al3+
and Fe3+) in the O sheet, and leaves a vacancy in the third octahedron; (ii) all the octahedral
sites are taken by divalent cations (usually Fe2+, Mg2+, Mn2+); and (iii) all the tetrahedra in
T sheet contains Si4+. A negatively charged layer case comes from (i) isomorphic
substitution of Al3+ for Si4+ in tetrahedral sites, and Mg2+ for Al3+ in octahedral sites; (ii)
the existence of vacancies. Layer charge is of great significance to 2:1 phyllosilicates,
because of which the exchangeable cations occupancy the interlayer space, and the
negative layer charge varies from 0.2 to 2.0 per formula unit for different phyllosilicates.
19
For 1:1 type phyllosilicates, the layer charge is usually zero per formula unit,
approximately24.
The surfaces of a clay mineral can be negatively, positively charged, or electrical neutral
due to different charging mechanisms on distinct surfaces. For basal surface, the prevailing
charging mechanism is isomorphic substitution, while for edge surface, surface charges
mainly come from the hydrolysis reactions of broken bonds28.
When immersed in an aqueous environment, metal ions will be adsorbed on the mineral
surfaces due to: (i) electrostatic interaction between metal cations and the negatively
charged or neutral mineral surfaces; and (ii) specific adsorption of divalent cations and
surface complexation.
2.2.3.3 Importance of anisotropic surface property
The layered structure of clay mineral makes the particle have distinct surfaces and thus,
anisotropic surface property for each surface. Anisotropic surface property dominates the
chemical and physical characteristics of clay minerals, such as colloidal stability, swelling,
ion exchange, etc28. For example, in the dewatering of oil sands tailings, the maximum
coagulation should occur at the point of zero charge (PZC) of the particles dispersed in the
suspension, however, the PZC obtained from the correlation used to calculate the
maximum coagulation point is only valid for isotropic particles, therefore, large
discrepancy exists between the calculated PZC and the optimal pH condition for particle
coagulation due to the anisotropic surface property of clay minerals. In this regard, the
probing of the anisotropic surface properties of phyllosilicates is critical to the development
of clay science and the applications of clay minerals in various fields.
20
2.2.4 Investigation techniques for surface property of clay minerals
The surface property of phyllosilicates is of significant importance and has been
extensively studied using various techniques, which are mainly electrophoretic mobility
(EPM) measurement, potentiometric titration, sum frequency generation spectroscopy and
AFM. However, due to the limitation of individual technique, it is not surprising to see
very distinct results for the same kind of clay mineral. Here we briefly introduce the pros
and cons of commonly used investigation methods.
2.2.4.1 Electrophoretic mobility (EPM) measurement
Electrophoretic mobility measurement is one of the common methods used to yield the zeta
potential of the particles when they are moving in an aqueous solution under the influence
of an electric field. Sondi et al.29 used the measurement of EPM to study the electrokinetic
behavior of three pure clay minerals: montmorillonite, illite and chlorites, and the
isoelectric point (IEP) of chlorite was found to be at pH 5.0 0.2, while there were no IEP
found for illite and montmorillonite. However, this method is used to deduce the zeta
potential of the whole particle with assumption that the particles are spherical, which is far
from the fact that phyllosilicates are platy and have anisotropic surface charging
characteristics.
2.2.4.2 Potentiometric titration
The principle of potentiometric titration method is based on the ion exchange in solution,
thus the results are not affected by the shape of the particles, but it still does not take the
anisotropic property of clay minerals into consideration. Also, this method is limited to
21
systems without specific adsorbing ions in the solution. Therefore, the PZC of clay
minerals determined by this method are not in good consistency30.
2.2.4.3 Sum frequency generation (SFG) spectroscopy
SFG is used to explore the surface property of individual surface of clay minerals. In SFG
method, two laser beams mix at an interface and generate an output beam. The frequency
of the output is equal to the sum of the two input frequencies, and the direction of the output
beam is given by the sum of the wavevectors of the two incident beams. When mineral
particles are in contact of water at different pHs, protonation/deprotonation reactions
happen and generate different interfacial water species. By analyzing the frequency
intensities and the orientations of different interfacial water species at different pHs, the
PZC of mineral particles can be estimated31. However, it is still a nascent technique without
routine experimental setup and interpretation of the results32, and there are few reports on
the investigation of phyllosilicate surface properties using this method28.
2.2.4.5 Atomic force microscopy
The atomic force microscope was invented by Gerber et al.33 in the 1980s. AFM is a
powerful and mature tool that can be used to image the topography of conducting and
insulating surfaces with high resolution, and even with atomic resolution in some cases. In
addition, AFM also allows the measurement of force-versus-distance curves, which can
give us valuable information, such as the elasticity, hardness, surface potential, surface
charge density, Hamaker constant etc. The single molecule force measurement of AFM,
which mainly involves the rupture of single chemical bonds and the stretching of polymer
chains, enables us to acquire the interactions of a single molecular pair34.
22
The commercially fabricated tips used to scan the surfaces of clay minerals are silicon
nitride (Si3N4) tip and silicon tip (Si), and the radii of the tips are usually between a few
nanometers and a few tens of micrometers. In addition, the force measurement can be as
sensitive as in pico-newton (pN) scale. So, AFM is very suitable for probing the surface
property of clay minerals (whose size is usually in the scale of micron), and the interaction
forces between AFM tip and clay surfaces in an aqueous/colloidal environment.
2.2.5 Probing anisotropic surface property of clay minerals using AFM
The use of AFM for surface property exploration of clay minerals was confined due to the
challenge of creating smooth enough surface. In recent years, due to the development of
ultramicrotome-cutting technology, which is able to provide surface smooth enough for
AFM force measurement between AFM tip and the clay surface, AFM has been used in
the study of the anisotropic surface properties of phyllosilicates, such as kaolinite, chlorite,
mica, talc, MoS2, illite, smectite, etc.
2.2.5.1 Sample preparation
When applying the AFM technique to determine the surface property of phyllosilicates,
sample preparation is a critical step because a selected surface is hard to obtain and surface
roughness greatly affects the accuracy of the results. A routine procedure has already been
developed to make selected basal plane and edge surfaces.
Kaolinite, which is composed of one silica tetrahedral basal plane, one alumina octahedral
basal plane, and edge surfaces, has been studied by Gupta35, Liu36, et al. The silica basal
surface is believed to be negatively charged permanently, while the alumina basal surface
is slightly pH-dependent. According to these charging characteristics, selected basal
23
surface can be prepared. In Gupta’s work35, silica basal surfaces were exposed by
depositing kaolinite particles onto glass substrates (which carries negative charges at pH >
4, thus the positively charged alumina surfaces preferentially attached to the negatively
charged glass substrate), and the alumina basal surfaces were exposed by depositing
kaolinite particles on a fused alumina substrate (which carries positive charges at pH < 6).
Liu et al.36 completed the research on kaolinite by studying its edge surfaces through
sandwiching kaolinite particles in the center of the epoxy resin, and the epoxy block was
trimmed by a razor under optical microscope to obtain a prismatic sample with kaolinite
particles exposed, then the sample was mounted into the slot of an ultramicrotome for
cutting.
For talc, mica, and chlorite, which are easily cleaved phyllosilicates, cleaving these
materials can expose atomically flat basal surfaces without contamination37. The
preparation of edge surfaces of these phyllosilicates could also follow the above-mentioned
experimental protocol for kaolinite edge surfaces.
2.2.5.2 Theoretical model
AFM force measurement between AFM tip and clay surfaces can be used to obtain the
surface charging characteristics of phyllosilicates under aqueous environment. In such
system, the interaction forces between phyllosilicate particle and AFM tip are governed by
the sum of van der Waals (vdw) force and electrical double layer (EDL) force, which is the
DLVO theory proposed by Derjaguin and Landau38 and Verwey and Overbeek39.
𝐹𝐷𝐿𝑉𝑂 = 𝐹𝑣𝑑𝑤 + 𝐹𝑒𝑑𝑙 (2.1)
24
DLVO theory is commonly adopted to fit the force curves obtained from AFM
measurement, and the derivation of the DLVO theoretical model for a pyramidal-shaped
tip and a flat substrate has been conducted by Drelich et al.40, which will be detailed
explained in Chapter 3.
2.2.5.3 Reported results
Since the application of AFM technique has been used in the research of anisotropic surface
properties of clay minerals, lots of the former results obtained from electrophoresis,
titration have been shown to be not accurate enough. The more accurate surface charging
characteristics of clay minerals should be achieved through this cutting-edge technique -
AFM. The results are collected and listed in Table 2.4.
One thing that needs to be addressed here is the definition and difference between PZC and
IEP, as they are frequently seen in reports to describe the zero charge characteristics of clay
mineral surface, but which are also not clearly separated. The PZC is defined as the point
at which the surface charge equals zero, and the IEP refers to the point at which the
electrokinetic potential (or zeta potential) equals zero. More detailed difference between
PZC and IEP was discussed by Kosmulski41.
25
Table 2.4 Anisotropic surface properties of clay minerals studied by AFM
Group Clay mineral
Cleavage
(Basal)
surface
Zero charge
condition for
basal surface
Zero charge
condition for
edge surface
Background
electrolyte
solution
Kaolin Kaolinite35,36
T: Si-O-Si IEP < pH 4,
pH-dependent
PZC < pH 4,
pH-dependent
Basal
surface: 1
mM KCl
solution
Edge
surface: 5
mM KCl
solution
O: Al-OH
pH 6 < IEP <
pH 8,
pH-dependent
Talc Talc42 T: SiOSi
siloxane
~ 30 mV,
pH-independent
PZC ~ pH 8,
pH-dependent
1 mM KCl
solution
Mica
Muscovite42 T: SiOSi
siloxane
~ 70 mV,
pH-independent
pH 7 < PZC <
pH 8,
pH-dependent
1 mM KCl
solution
Illite43 T: SiOSi
siloxane
−18mV at pH
2.9, −37mV at
pH 8.3,
pH-dependent
- 1 mM KCl
solution
Chlorite Chlorite44
T: SiOSi
siloxane
IEP < pH 5.6,
pH-independent IEP ~ pH 8.5,
strongly
pH-dependent
1 mM KCl
solution O: MgOH
hydroxyl
IEP > pH 9,
slightly
pH-dependent
Sulfide
mineral Molybdenite SMoS
None detected
in pH 3.0-11.1
PZC ~ pH 3,
pH-dependent
10 mM
NaCl
solution
2.2.6 Influence of anisotropic surface property
2.2.6.1 Influence on processing of clay minerals
Mineral processing is traditionally regarded as the processing of ores or other materials to
obtain concentrated products45. Naturally, clays are mixtures of clay minerals and non-clay
minerals, and in industry, extensive processing technologies are adopted to yield clay
26
minerals with a certain degree of purity. Typical processes for obtaining clay minerals may
involve24:
i) purification;
ii) wet/dry processing;
iii) particle size separation;
iv) bleaching and magnetic separation;
v) flotation and selective flocculation;
vi) drying and calcination;
vii) chemical modification and activation.
The study on anisotropic surface properties of clay minerals can help us collect useful
information, such as the clay mineral topography, surface charging characteristics, the
suspension rheology46, etc., which would be beneficial to clay mineral processing in
various ways, for example24:
1) knowledge of the surface charging characteristics is useful in setting the optimal pH
environment for the preparation of a stable dispersion of clays in wet processing, thus
avoiding the aggregation of clay mineral particles;
2) rheology properties of soda-activated bentonites and the layer arrangement within
individual montmorillonite particles determine the optimum addition of soda in the
processing of raw bentonites;
27
3) particle shape and surface charge of clay minerals are important properties that affect
the sedimentation or filtration process, which can be used to choose effective process aid
agents.
2.2.6.2 Influence on industrial applications
As mentioned above, clay minerals have widely been used in numerous industrial
applications. Anisotropic surface property of clay minerals will also affect their
applications, such as:
1) in oil sands extraction, the understanding of the anisotropic surface-chemistry properties
can guide the development of flotation reagent in the flotation separation process47;
2) in oil sands tailings management, surface charging characteristics of kaolinite, chlorite,
illite, etc. are critical in finding suitable coagulants or flocculants to accelerate the
dewatering process35,36,48;
3) in flotation process of molybdenite, the anisotropic surface properties can provide
explanation of the variation in flotation recovery over the change of particle size49.
2.3 Kaolinite
2.3.1 Background information for kaolinite
Kaolinite (Al2[Si2O5](OH)4) is a kind of phyllosilicates (sheet silicates), which accounts
for about 69% of the clay in oil sands open pit mining, and other clay minerals include 28%
illite, 0.3% smectite, and 1% chlorite, and 1.7% mixed layer clay, as shown in Figure 2.6.
This overwhelming majority proportion of kaolinite makes it significantly important to
28
deeply and thoroughly understand its crystal structure and surface properties, which govern
the wettability, aggregation, dispersion and flotation of kaolinite50,51. Therefore, the
aggregation of kaolinite particles could be manipulated so as to benefit the treatment of oil
sands tailings.
Figure 2.6 Relative amounts of clay minerals in oil sands deposits52.
2.3.2 Structure of kaolinite
Kaolinite is usually white and soft with a hardness of 2-2.553. Kaolinite has a perfect two-
layer clay as shown in Figure 2.7, which is composed of three kinds of surfaces, one silica
tetrahedral basal plane, one alumina octahedral basal plane, and edge surfaces. The bilayers
are bonded together by hydrogen bonds between the hydroxyl ions of the octahedral sheet
and the oxygen atoms of the silica tetrahedral sheet54. The non-expandable kaolinite has a
low isomorphous substitution (which leads to a relatively low permanent charge) and a low
cation exchange capacity (CEC).
29
Figure 2.7 Structure of kaolinite clays36. Red: oxygen; yellow: silicon; purple: aluminum;
white: hydrogen.
The different surfaces of kaolinite have different charging characteristics, which is
governed by distinct charging mechanisms: isomorphous substitution on the basal plane
and hydrolysis reactions of broken bonds on the edge surface24.
In aqueous solution, the departure of the compensating cations results in the appearance of
a surface charge on the external basal surfaces and near the layer edges of kaolinite. The
silica basal plane is negatively charged permanently, and pH independent, while the
alumina basal plane is slightly pH-dependent. The exposed aluminum hydroxyl surface can
be protonated at low pH and deprotonated at high pH by following hydrolysis reactions, so
the Al-OH surface is positively charged at low pH and negatively charged at high pH55.
Whether the edges carry positive or negative charges strongly depends on pH and Si-Al
ratio, so the PZC varies24. PZC is an important parameter because particles carry a net
negative charge when the pH is above PZC, otherwise, particles carry a net positive surface
charge while below PZC.
30
Table 2.5 Summary of reported PZC and IEP for kaolinite.
PZC Measurement
method IEP
Measurement
method
6-6.555 Potentiometric
titration <2.456
ζ-potential
measurements
~4.557 Titration <327 Electrophoresis
measurements
<436 AFM 2. 958 ζ-potential
measurements
The IEP reported are consistent. However, the indirect derivations of the surface charges
of kaolinite particle vary over a wide range, and some of them can even be very misleading.
Therefore, a reliable experimental measurement of the charging characteristics of kaolinite
surfaces is necessary53.
2.3.3 Particle aggregation
In hydrous dispersions, the dispersed clay particles can be observed to collide frequently
due to the Brownian motion, but they separate again after the collision. However, this
situation can be changed by just adding a small amount of salt. The clay particles stick
together upon collision, and this phenomenon is called flocculation or coagulation52. The
generation of this phenomenon can be explained by the theory of the stability and
flocculation of colloidal dispersions. The attraction between dispersed particles is because
of the van der Waals attraction force, while the repulsive force comes from the electrical
nature. The particle charge can be altered by the presence or absence of ionized salt in the
solutions, and the phenomenon mentioned above is because repulsion decreases when the
ion concentration increases.
31
The interaction between kaolinite particles are governed by van der Waals force, electrical
double layer force, hydrophobic force, and hydration force, etc., and according to the three
different surfaces of kaolinite, there are three possible kinds of particle association modes:
F(face) to F, E (edge) to F, and E to E. As show in Figure 2.8, a card-house structure
was proposed for the kaolinite particle aggregates in suspension26.
Figure 2.8 Stairstep card-house structure of aggregated kaolinite particles53.
AFM is capable of measuring the above mentioned long-range and short-range interaction
forces between particles, and of characterizing the surfaces for microscopic clay particles.
Combined with DLVO theory, the analysis of AFM-measured forces can be used to
compare the experimental results with the theoretical model predictions, and derive the
surface potentials/charge densities, which are very useful for predicting the particle
aggregation formation and thus guiding the dewatering process in oil sands tailings
treatment40.
2.4 Ion distribution at a solid-liquid interface
Generally, an aqueous environment containing various ions is inevitable when utilizing or
removing clay mineral particles, and therefore the importance of studying ion adsorptions
and ion distributions at the charged clay-liquid interface is self-evident.
32
2.4.1 Electric double layer theory
The study of charged interfaces can be traced back to about two hundred years ago. The
first theoretical model for the charge distribution at the electrified interface was brought up
by Helmholtz59, which is the simplest model of electric double layer (EDL). The Helmholtz
model treated the charge distribution on a solid surface as a plane layer of counter ions
absorbed on the surface and neutralized the surface charges, which was unrealistic
especially for solid-liquid interface. Gouy60 and Chapman61 developed the Helmholtz
model by taking into account the diffusion and thermal motion of ions. In 1924, Stern62
divided the EDL into an inner layer, the Stern layer, and an outer layer, the diffuser layer
or Gouy layer. The Stern layer starts from the inner Helmholtz plane (IHP) where the
specific adsorption occurs, and ends at the outer Helmholtz plane (OHP) where the non-
specific adsorption of hydrated counterions starts. The Stern model can give a good account
of the ion adsorption behavior at the solid-liquid interface in dilute electrolyte solutions63.
For decades, the theoretical EDL models and corresponding calculations and experiments
served for solving a universal question: how and where the ions adsorb on charged surfaces
in liquid? Different approaches led to inconsistent conclusions.
2.4.2 Debate on ion adsorption mechanisms
Although there was consensus that monovalent ions (for example, Na+, K+, Cl-) were
“indifferent” ions because they could not change the point of zero charge (PZC) of oxide
surfaces, there was a debate on whether monovalent ions could adsorb at the Stern layer.
Some believed that monovalent cations were strongly hydrated and could only interact with
the surface beyond the OHP, whilst monovalent anions behaved like individual ions and
could adsorb specifically at the interface64 because they were less hydrated. Another
33
explanation to the unchanging PZC of metal oxide in simple electrolyte solutions was equal
adsorption of monovalent cations and anions in the inner layer at the PZC65.
Divalent cations were regarded as specifically adsorbed at the solid-liquid interface at all
times65,66. When Ardizzone et al.66 observed the cadmium ions and lead ions started to
strongly adsorbed onto the hematite surface at pH above 7 and above 4, respectively, the
high adsorption value ascribed to bare ions alone seemed implausible. James and Healy67–
69 attributed the specific adsorption to the partially hydrolyzed species of divalent cations
(M(OH)+), and they believed that divalent metal ions had little contributions to the
adsorption in the inner region of electrical double layers.
2.4.3 Atomic resolution imaging
The identification of specific adsorption has been ambiguous because previous
experiments were mostly indirect (for example, depending on the examination of PZC) and
corresponding analysis or calculations had to rely on the model of EDLs. All models were
based on various levels of assumptions with specific constrains. Therefore, a completely
model-independent method to identify specific adsorption at solid-liquid interface has been
imperative.
AFM is a powerful tool to study the surface morphology, which has been successfully used
to provide atomic resolution imaging on synthesized and natural clay particles. Gan et
al.70,71 obtained hexagonal lattice arrangement with a periodicity of about 0.47 nm of
sapphire (0001) surface. Mugele and Siretanu et al.72,73 studied ion adsorption on perfectly-
cleaved mica and synthesized gibbsite surfaces, and they observed the surface structure of
mica changed from hexagonal rings to rectangular symmetry due to high concentration of
34
divalent cation adsorption. Mugele and Siretanu et al. 74 went further to explore the atomic
structure and surface defects of natural clay mineral, such as kaolinite, montmorillonite,
etc.. In this work, we used the technique of atomic resolution imaging to compare the
surface lattice structure before and after monovalent and divalent cations addition at low
and high solution pHs, hoping to prove the ion adsorption mechanisms we proposed during
the exploration of the influence of cations on the Stern potentials of different kaolinite
surfaces.
35
Chapter 3 Materials and Methods
3.1 Materials
20 40 60 80 100
-100
0
100
200
300
400
500
600
700
800
(1 3
1)
(-2
0 1
)
(1 1
1)
(1 -
3 2
)
(-2
0 4
)
(-3
-3
1)
(2 0
0)(-1
1 0
)
Inte
nsity (
AU
)
(degrees)
(0 0
1)
(0 0
2)
(0 2
0)
(1 -
3 1
)
(-1
-1
1)
Figure 3.1 X-ray diffraction (XRD) analysis of kaolinite.
Preparation of kaolinite suspension: kaolinite particles used in this research were obtained
from IMERYS kaolinite, Inc., and were used as received. X-ray diffraction (Rigaku XRD
Ultima IV, Rigaku Corporation) pattern of the kaolinite sample shown in Figure 3.1 fits
very well with XRD pattern of 80-0886, which represents kaolinite with chemical formula
of Al2Si2O5(OH)4. Original kaolinite suspension (3.0 wt%) was prepared in high purity
Milli-Q water (Millipore Inc.) with a resistivity of 18.2 MΩ∙cm-1. The suspension was
magnetically stirred for 20 min, and then adjusted to pH 9 using 0.1 M HCl and 0.1 M
KOH solutions, followed by five minutes of sonication to fully disperse the particles. After
two hours of gravity settling (or four hours for edge surfaces preparation), the sediment of
the suspension was discarded to remove the large particles, and the supernatant was
36
centrifuged at 500 g for 30 min in an Ultra Centrifuge (Sorvall WX80, Thermo Scientific
Inc.) to remove the super-fine particles. The sediment was collected and centrifuged at 300
g for another 30 min. The supernatant was again decanted and the centrifuge cake was
diluted in Milli-Q water (to about 3 wt% of clay) and then sonicated for 20 min to be fully
dispersed. The re-dispersed kaolinite suspension was adjusted to about pH 5 for kaolinite
basal planes preparation, and to pH 9 for kaolinite edge surfaces preparation.
The average particle size of the original kaolinite suspension was determined using
MasterSizer 3000 (Malvern Instruments Ltd., UK) to be 6.6 µm. The re-dispersed kaolinite
particles have the average size of 868 nm and 660 nm for two-hour and four-hour gravity
settling, respectively, which were measured by a ZetaSizer Nano (Nano ZS, Malvern
Instruments Ltd., UK).
3.2 Sample preparation
The sample preparation procedure developed by Gupta et al.35 and the ultramicrotome
cutting technique adopted by Yan et al.42 enabled us to expose the two basal planes and the
edge surface of kaolinite, respectively, to investigate the surface charging properties. At
certain pH, the siloxane basal plane of kaolinite (referred to as Si-basal plane) is negatively
charged while the aluminum oxy-hydroxyl basal plane of kaolinite (referred to as Al-basal
plane) is positively charged. By using this property, Gupta et al.35 successfully exposed the
Si-basal plane and Al-basal plane by depositing kaolinite suspension onto a negatively
charged glass substrate and a positively charged alumina substrate, respectively. To expose
the edge surfaces, Yan et al.42 embedded a thin piece of phyllosilicate clay in a resin block
37
and cut the cross section using an ultramicrotome. The schematics of the preparation of
kaolinite basal planes and edge surfaces can be seen in Figure 3.2.
Figure 3.2 Schematics of preparing kaolinite (a) Si-basal plane, (b) Al-basal plane, and (c)
edge surfaces.
3.2.1 Kaolinite basal planes
The microscopic glass slides (Fisher Scientific Inc.), the silica wafer (NanoFab, University
of Alberta, CA) and the sapphire wafer (C-plane, (0001), University Wafer Inc.) were cut
into 10-by-10-mm pieces. The glass slides, sapphire wafers and silica wafers were cleaned
with piranha solution, composed of three volumes of sulfuric acid (H2SO4 96%) and one
volume of hydrogen peroxide (H2O2 30%), for 30 min, followed by rinsing with large
amounts of Milli-Q water, sonicating for 15 minutes in Milli-Q water, and blowing dry
with ultra-high purity N2 gas. It was expected that this process would remove the organic
and inorganic contaminations on the substrates.
Two drops of the kaolinite suspension at pH 5 were put on the clean glass and sapphire
substrates, which were placed on a ~70˚C hot plate to dry. The purpose of this technique
was to expose the kaolinite Si-basal plane and Al-basal plane, respectively.
38
3.2.2 Kaolinite edge surfaces
The re-dispersed kaolinite suspension was pipetted onto a layer of hardened epoxy resin
(Electron Microscopy Sciences, Hatfield, PA, USA), which was placed on a hot plate at
about 120 ˚C. The kaolinite particles were expected to flatly and evenly spread over the
resin surface in such way. Another layer of the epoxy resin was put on top of the dried
kaolinite particles. The two layers of the epoxy resin and the kaolinite particles in the center
together, formed a resin block with a sandwich structure. The resin block was trimmed by
a trimmer (Leica EM TRIM2, Leica Microsystems Inc.) to create a tetragonal top surface
and four trapezoid side faces as shown in Figure 3.2(c). The tetragonal top surface should
be as small as possible to reduce the friction generated during the ultramicrotome cutting
procedure. The trimmed resin block was then mounted as perpendicular as possible on the
ultramicrotome (Leica EM UC7, Leica Microsystems Inc.) for cutting. After cutting with
a diamond knife on the ultramicrotome, the resin block was rinsed with Milli-Q water and
blown dry with high-pressure ultrapure N2 gas to remove the remained flakes on the surface,
and it was then glued on a clean glass substrate. Before being used in AFM force
measurement, the resin block was put in a UV-ozone for five minutes, rinsed with a large
amount of Milli-Q water, and blown dry with ultra-high purity N2 gas to remove the organic
contaminants.
3.3 Atomic force microscopy
3.3.1 AFM force measurement
An Atomic Force Microscope (Bruker Dimension Icon-PT, Bruker Corporation) was used
to measure the interaction forces between the AFM tip and the kaolinite surfaces. Silica
39
nitride tips (DNP-10, Bruker AFM Probes, Bruker Nano Inc.) with a tip radius of about 20
nm were used in the research of kaolinite basal planes. Because the width of the edge
surfaces is between 30 nm and 80 nm, silicon tips (SNL-10, Bruker AFM Probes, Bruker
Nano Inc.) with a smaller tip radius of about 2 nm were used to study kaolinite edge
surfaces. The spring constant of each cantilever was evaluated by the Thermal Tune
Function provided in the NanoScope Software V8.15 (Bruker Corporation). Dimensional
analysis of the probes was carried out on the SEM images of the probes.
The influence of pH on the Stern potentials of kaolinite surfaces was studied at pH 3, 5, 6,
8, and 10 in 10 mM KCl solutions. The effect of divalent cations were studied in 10 mM
KCl solutions with different concentrations of Ca2+ and Mg2+ at pH 8 and 11, respectively.
In 10 mM KCl solution, the thickness of the electrostatic double-layer (𝜅−1) is about 3.03
nm, which is much smaller than the thickness of kaolinite particle (see Figure 3.3), so the
tip-substrate electrostatic double-layer would not overlap with the tip-kaolinite surface
electrostatic double-layer75. When divalent cations were added into the 10 mM KCl
background solutions, the thickness of the electrostatic double-layer would be compressed
even more.
Figure 3.3 Illustration of the electric double layer (shown as grey thin layer) surrounding
the AFM tip, kaolinite and substrate in electrolyte solutions with 𝜅−1 less than the
thickness (d) of the kaolinite particle.
40
Force measurements between the AFM tip and the sample surface were performed after
the tip-solution-sample surface system was stabilized for 30 min. The topography image
of the kaolinite surfaces was scanned and captured under contact mode by the Point and
Shoot Function provided in the NanoScope Software V8.15, and then the interaction force
curves were taken 5 times at each chosen location on this image. For each background
solution condition, more than 500 force curves were taken at 3 ~ 4 different places on every
sample to guarantee the reproducibility of the data. The measured force curves were fitted
with theoretical DLVO force curves to derive the Stern potentials of kaolinite surfaces.
The raw force curves were Deflection Error-Z distance data, which were converted to
Force-Separation distance curves by NanoScope Analysis Version 1.8 (Bruker
Corporation). Baseline corrections of the force curves were also conducted before they are
fitted with theoretical DLVO force curves.
3.3.2 AFM tip evaluation
The geometry of the AFM probes was obtained by analyzing the images of the probes (see
Figure 3.4), which were captured by a Scanning Electron Microscope (SEM)/Mineral
Liberation Analyzer (Quanta 250, FEI Company). The pyramid-shape probe was
composed of a spherical apex and a conical tip.
The curvature and the conical angle of the spherical apex of the tip were determined by
fitting it with a circle. The curvatures for Si3N4 tip and Si tip were determined to be about
60 nm and 30 nm, respectively. The conical angle (2β) of the Si3N4 tip and Si tip were
around 40˚ and 35˚, respectively. SEM images were taken to check the status of the probe.
41
If they showed that the probe was worn or damaged after the force measurement, the
experimental results obtained by this tip would not be used any further.
Figure 3.4 SEM images of the side view of (a) a cantilever with a DNP-10 AFM tip on top
at low magnification, (b) DNP-10 AFM tip at high magnification, (c) cantilever with a
SNL-10 AFM tip on top at low magnification, and (d) SNL-10 AFM tip at high
magnification.
3.3.3 Atomic resolution imaging by AFM
Atomic resolution imaging was performed using a Dimension ICON-ScanAsyst AFM
(Bruker Dimension Icon-PT, Bruker Corporation). Fastscan-C probes with a tip radius of
42
5 nm, a cantilever of 40 µm and a spring constant of 0.8 N/m were used. Kaolinite samples
were rinsed with Milli-Q water and blown dry with ultra-pure N2 gas to remove loosely
attached particles. Kaolinite sample was immersed in background electrolyte solutions for
30 min before performing the AFM imaging. ScanAsyst mode with auto-control was first
used to find a kaolinite particle firmly attached to the substrate. The tip was then withdrawn
from the sample surface. After turning off the auto-control, the peak force engage setpoint
was set as 0.04 V and the tip was engaged in contact with the kaolinite particle surface
again. To improve the quality of images, scanning parameters were carefully adjusted after
zoom-in on a kaolinite particle. To reduce the drift, the scan rate was increased to 2 Hz for
512 (Samples/Line) × 512 (Lines) data format. The obtained images were treated using
NanoScope Analysis Version 1.8 (Bruker Corporation), and the image processing
procedure was provided in Appendix A.
3.4 Theoretical model
Figure 3.5 Schematic illustration of the AFM tip-flat surface system used in the theoretical
calculation. Angles α and β are the geometrical angles for the spherical cap at the tip apex
and the conical tip body, respectively, and α + β = 90°; D refers to the separation between
43
the end of the tip apex and the surface of the substrate; L is the distance from the AFM tip
to the substrate; R is the radius of the spherical cap at the tip apex; r is the radius of the tip
at a given vertical position40.
The geometry of the pyramidal-shape tips shown in Figure 3.5 can be reasonably
approximated as a cone with a spherical cap at its apex as illustrated in Figure 3.5. The
interaction force between an AFM tip and a flat sample surface is dominated by van der
Waals force and electrostatic double-layer force, which is described by the classical DLVO
theory39.
The derivation of the DLVO theory that can be applicable to the AFM tip-flat substrate
system was reported in the literature40,49 and only the final equations are listed below:
𝐹𝐷𝐿𝑉𝑂 = 𝐹𝑣𝑑𝑤 + 𝐹𝑒𝑑𝑙 = 𝐹𝑇𝑆𝑆 + 𝐹𝑇𝑆
𝐶 (3.1)
van der Waals force:
𝐹𝑣𝑑𝑤 =𝐴
6[
(𝑅+𝐷)−2𝐿1
𝐿12 −
𝑅−𝐷
𝐷2 ] −𝐴
3 tan2 𝛼(
1.0
𝐿1+
𝑅 sin 𝛼 tan 𝛼−𝐷−𝑅(1−cos 𝛼)
𝐿12 ) (3.2)
where 𝐴 is the combined Hamaker constant76 for the tip-solution-substrate system. Based
on the Lifshitz theory77, the nonretarded Hamaker constant for two macroscopic phases 1
and 2 interacting across a medium 3 is expressed by78:
𝐴 =3
4𝑘𝐵𝑇 (
𝜀1−𝜀3
𝜀1+𝜀3) (
𝜀2−𝜀3
𝜀2+𝜀3) +
3ℎ𝜈𝑒
8√2
(𝑛12−𝑛3
2)(𝑛22−𝑛3
2)
(𝑛12+𝑛3
2)1 2⁄ (𝑛22+𝑛3
2)1 2⁄ (𝑛12+𝑛3
2)1 2⁄ +(𝑛22+𝑛3
2)1 2⁄ (3.3)
where 𝜀1, 𝜀2, and 𝜀3 are the static dielectric constants of the three media; 𝑛1, 𝑛2, and 𝑛3
are the refractive indexes for three media; 𝑘𝐵 is the Boltzmann constant, 1.38×10-23J/K; T
44
is the absolute temperature in Kelvin; ℎ is the Planck’s constant, 6.626×10−34 J·s; 𝜈𝑒 is the
main electronic absorption frequency in the UV region, which is assumed to be the same
(3.0 × 1015s−1) for all materials in this work.
Table 3.1 Values used to calculate the Hamaker constants.
Materials
Static Dielectric
Constant Refractive Index Hamaker Constant
ε n A
Silica79 3.82 1.448
Water79 78.5 1.333
Silicon nitride79 7.4 1.988
Alumina78 10.1-11.6 1.753
Silicon78 11.6 3.44
Silicon nitride-
water-silica wafer 2.01 × 10−20J
Silicon nitride-
water-silica basal
plane of kaolinite
2.01 × 10−20J
Silicon nitride-
water-alumina basal
plane of kaolinite
6.33 × 10−20J
Silicon-water-silica
wafer 4.33 × 10−20J
Silicon (𝐴11)36 1.36 × 10−19
Kaolinite edge
surface (𝐴22)57 1.20 × 10−19
Water (𝐴33)78 3.70 × 10−20
Silicon-water-edge
surface of kaolinite 2.72 × 10−20J
45
Due to the lack of the information of the dielectric constant and refractive index of kaolinite
edge surface, the combined Hamaker constant for silicon-water-kaolinite edge surface was
estimated by the following equation78:
𝐴 = (√𝐴11 − √𝐴33) × (√𝐴22 − √𝐴33) (3.4)
where, A11, A22 and A33 refer to the Hamaker constants of individual medium 1, 2 and 3,
respectively. In this case, A11, A22 and A33 are the Hamaker constants of silicon, kaolinite
edge surface and water, respectively.
All the values needed by the calculation of the Hamaker constants are listed in the above
Table 3.1.
Electrostatic double-layer force:
For the case of constant surface potential boundary condition (BC):
𝐹𝑒𝑑𝑙 = 4𝜋𝜀0𝜀𝜓𝑇𝜓𝑆(𝑎0𝑒−𝜅𝐷 − 𝑎1𝑒−𝜅𝐿1) − 2𝜋𝜀0𝜀(𝜓𝑇2 + 𝜓𝑆
2)(𝑎2𝑒−2𝜅𝐷 − 𝑎3𝑒−2𝜅𝐿1) +
4𝜋𝜀0𝜀𝜅
𝑡𝑎𝑛 𝛼∙ 𝑏1𝜓𝑇𝜓𝑆𝑒−𝜅𝐿1 − 𝑏2
(𝜓𝑇2 +𝜓𝑆
2)
2𝑒−2𝜅𝐿1 (3.5)
For the case of constant surface charge density BC:
𝐹𝑒𝑑𝑙 =4𝜋
𝜀0𝜀𝜅2𝜎𝑇𝜎𝑆(𝑎0𝑒−𝜅𝐷 − 𝑎1𝑒−𝜅𝐿1) +
2𝜋
𝜀0𝜀𝜅2(𝜎𝑇
2 + 𝜎𝑆2)(𝑎2𝑒−2𝜅𝐷 − 𝑎3𝑒−2𝜅𝐿1) +
4𝜋
𝜀0𝜀𝜅 tan 𝛼∙ 𝑏1σ𝑇σ𝑆𝑒−𝜅𝐿1 + 𝑏2
(σ𝑇2 +σ𝑆
2)
2𝑒−2𝜅𝐿1 (3.6)
46
Theoretically, the electrostatic double-layer force was derived under either the constant
surface potential or the constant surface charge density BC. However, in reality, especially
at a short distance, the electrical double-layer force would lie between the two conditions28.
For the case that one surface is under constant surface potential condition and the other is
under constant surface charge density condition, the mixed BC should be adopted to fit the
measured force curves, and the equations49 for this case are given as follows:
𝑊𝑒𝑑𝑙𝑆 = 𝜋𝑅 [2𝜓𝑠𝜎𝑇 ∫ sech(𝜅𝐿)𝑑𝐿 + (
𝜎𝑇2
𝜀𝜀0𝜅− 𝜀𝜀0𝜅𝜓𝑠
2) × ∫ (tanh(𝜅𝐿) − 1)𝑑𝐿𝐿1
𝐷
𝐿1
𝐷] (3.7)
𝑊𝑒𝑑𝑙𝐶 =
𝜋𝑏3
tan 𝛼[2𝜓𝑠𝜎𝑇 ∫ sech(𝜅𝐿)𝑑𝐿 + (
𝜎𝑇2
𝜀𝜀0𝜅− 𝜀𝜀0𝜅𝜓𝑠
2) ∫ (tanh(𝜅𝐿) − 1)𝑑𝐿∞
𝐿1
∞
𝐿1] +
𝜋
tan 𝛼[2𝜓𝑠𝜎𝑇 ∫ sech(𝜅𝐿)𝐿𝑑𝐿 + (
𝜎𝑇2
𝜀𝜀0𝜅− 𝜀𝜀0𝜅𝜓𝑠
2) ∫ (tanh(𝜅𝐿) − 1)𝐿𝑑𝐿∞
𝐿1
∞
𝐿1] (3.8)
𝑊𝑒𝑑𝑙𝜓−𝜎
= 𝑊𝑒𝑑𝑙𝑆 + 𝑊𝑒𝑑𝑙
𝐶 (3.9)
𝐹𝑒𝑑𝑙𝜓−𝜎
= − (𝜕𝑊𝑒𝑑𝑙
𝜓−𝜎
𝜕𝐷) (3.10)
where,
𝐿1 = 𝐷 + 𝑅(1 − cos 𝛼) , 𝑎0 = 𝜅𝑅 − 1 , 𝑎1 = 𝜅𝑅 cos 𝛼 − 1 , 𝑎2 = 𝑎0 + 0.5 , and 𝑎3 =
𝑎1 + 0.5
𝑏1 = [𝑅 sin 𝛼 −𝐷+𝑅(1−cos 𝛼)
tan 𝛼] +
1
tan 𝛼∙ [(𝐿1 +
1
𝜅)]
𝑏2 = [𝑅 sin 𝛼 −𝐷+𝑅(1−cos 𝛼)
tan 𝛼] +
1
tan 𝛼∙ [(𝐿1 +
1
2𝜅)]
𝑏3 = 𝑅 sin 𝛼 tan 𝛼 − 𝐿1
47
Here, superscripts S and C refer to the spherical cap of the tip apex and the conical tip body,
respectively; subscripts S and T refer to the substrate and the AFM tip, respectively; 𝜀 and
𝜀0 are dielectric constant of the medium and the permittivity of vacuum, respectively; e is
the electronic charge, 1.602 × 10−19C; 𝜓 is the surface potential; 𝜎 is the surface charge
density; 𝜅 is the reciprocal of the Debye length80, which is a characteristic parameter
describing the thickness of EDL or decay of the electric potential:
𝜅−1 = (𝜀𝜀0𝑘𝐵𝑇
∑ 𝑛𝑖0𝑒2𝑧𝑖2
𝑖)
1/2
(3.11)
where 𝑛𝑖0 is bulk concentration of electrolyte i, ions/m3; 𝑧𝑖 is the valence of ion i.
The surface potential and the surface charge density are linked to each other by the
Grahame equation81:
σ = √8𝑐0𝜀𝜀0𝑘𝐵𝑇sinh (𝑒𝜓
2𝑘𝐵𝑇) (3.12)
where 𝑐0 is the bulk concentration of the salt; 𝑘𝐵 is the Boltzmann constant, 1.38×10-23J/K;
and T is the absolute temperature.
48
Chapter 4 Determination of Stern Potentials of Kaolinite
Surfaces: Effect of Divalent Cations and Solution pH
4.1 Introduction
In the processing of valuable resources, such as flotation or tailings treatment, divalent
cations should be removed to promote flotation efficiency or added to reclaim more process
water since they can alter the surface potentials of clay minerals. The concentration of
divalent cations is essential in their applications in industry. In real systems, a much lower
dosage of divalent cations than the critical coagulation concentration is enough to achieve
effective coagulation, as the specific adsorption nature of the corresponding monohydroxyl
ions would lead to charge reversal of clay particles. However, the impact of divalent cations
on the Stern potentials of kaolinite still lacks investigation. Therefore, in this work, we
focused on the anisotropic surface properties of kaolinite particles and took a deeper look
at the potential determining abilities of different divalent cations by AFM force
measurements. The interaction forces between the AFM tip and distinct kaolinite basal
planes in 10 mM KCl solutions containing Ca2+ and Mg2+ were fitted using DLVO theory
to derive the Stern potentials of different kaolinite surfaces. In addition, we explored the
mechanisms of divalent cations adsorbing on different kaolinite surfaces based on the
experimental results, aiming to provide clear insights into the colloidal behavior of
kaolinite that governs its vast applications in lab research and industry.
49
4.2 Results and discussion
4.2.1 AFM tip calibration
To obtain the Stern potential of selected kaolinite surface through the measured interaction
force between kaolinite surface and AFM tip, we first examined the Stern potential carried
by the AFM tip. Since the Stern potentials of silica wafer in 10 mM KCl solutions at
different pH environments were well studied and showed good consistency82–84, silica
wafer was thus used to obtain the Stern potential of the AFM tip. Stern potentials of silica
wafer in 10 mM KCl solutions with divalent cations have not been investigated. Therefore,
the zeta potentials of the silica wafer (see Figure 4.1) measured by SurPASS Electrokinetic
Analyzer (Anton Paar USA Inc.) using streaming potential method were adopted to
calibrate the AFM tips.
0 1 2 3 4 5
-80
-60
-40
-20
0
20
40
Ze
ta P
ote
ntia
l (m
V)
Concentrations of Divalent Cation (mM)
Ca2+
at pH 8
Mg2+
at pH 8
Ca2+
at pH 11
Mg2+
at pH 11
Figure 4.1 Zeta Potential of silica wafer in 10 mM KCl solutions at pH 8 and 11 with
varying concentrations of divalent cations (Ca2+ and Mg2+, respectively) determined using
the streaming potential method.
50
It can be seen in Figure 4.1 that the zeta potential of silica wafer is -52 mV in 10 mM KCl
solution at pH 8, which is slightly lower in magnitude than the Stern potential of silica
wafer (-59 mV) under this condition. This is because the Stern potential was measured at
a closer distance to the solid surface than the zeta potential. The zeta potential of the silica
wafer was determined to be -43 mV when 0.1 mM Ca2+ was added into the solution, and it
reached to -33 mV and -31 mV at 0.5 mM and 1 mM additions of Ca2+, respectively. The
zeta potential of silica wafer changed considerably to -20 mV when the concentration of
Ca2+ was increased to 5 mV. The zeta potentials of silica wafer in KCl solutions with the
addition of Mg2+ were found to be very close to those with the addition of Ca2+.
Yan et al.85 reported that the difference between the Stern potential and the zeta potential
became smaller when the electrostatic double-layer was further compressed by adding
higher concentrations of divalent cations. When 0.1, 0.5, 1 and 5 mM divalent cations were
added into the 10 mM KCl solution, the Debye length of a solid surface was compressed
from 3.03 nm to 2.99, 2.83, 2.66, and 1.92 nm, respectively. When the electrostatic double-
layer was further compressed by adding higher concentrations of divalent cations, the
difference between the Stern potential and the zeta potential became smaller. Therefore, it
is feasible to use the zeta potentials of silica wafer to evaluate the Stern potentials of the
AFM tips when divalent cations are added to the KCl solution.
Their conclusion was also confirmed by comparing the measured zeta potentials (using
ZetaSizer Nano) of Si3N4 nano particles with the Stern potentials of AFM Si3N4 tip
calibrated by AFM force measurement using silica wafer (see Figure 4.2).
51
0 1 2 3 4 5
-100
-80
-60
-40
-20
0
20
40
Ele
ctr
ica
l P
ote
ntia
l (m
V)
Concentration of Ca2+
(mM)
Zeta Potential at pH 8
Zeta Potential at pH 11
Stern Potential at pH 8
Stern Potential at pH 11
(a)
0 1 2 3 4 5
-100
-80
-60
-40
-20
0
20
40
(b)
Ele
ctr
ica
l P
ote
ntia
l (m
V)
Concentration of Mg2+
(mM)
Zeta Potential at pH 8
Zeta Potential at pH 11
Stern Potential at pH 8
Stern Potential at pH 11
Figure 4.2 Comparison of the zeta potential of Si3N4 nano particles with the Stern potential
of AFM Si3N4 tip in 10 mM KCl solutions at pH 8 and 11 containing different
concentrations of (a) Ca2+ and (b) Mg2+.
52
In this research, pH 3 was used instead of pH 4 to avoid the PZC of silica nitride tip, which
was reported86 to be near pH 4. The measured force curves of the interaction between the
AFM tips and the silica wafer (see Figure 4.3 and Figure 4.4 for silicon nitride tip and
silicon tip, respectively) were fitted with the theoretical force curves calculated based on
the classical DLVO theory, which were computed using MATLAB (The MathWorks Inc.),
to derive the Stern potentials of the AFM tips. The experimental data fitted well with the
theoretical data at the separation distance > 3 nm, but they deviated at a very short
separation. This phenomenon was also observed in the investigation of the charging
characteristics of different kaolinite surfaces, and it is probably a consequence of the
surface roughness and the repulsive, short-distance and non-DLVO hydration force
between the AFM tip and the substrate/sample surface, which was not taken into
consideration in this study. More details about the error analysis can be found in Appendix
B. Therefore, the experimental force curves were always higher than the corresponding
DLVO force curves when the separation was less than 3 nm. Also, the “jump-to-contact”
point can be observed at the separation distance < 2 nm.
53
Figure 4.3 Interaction forces between the AFM Si3N4 tip and the silica wafer in 10 mM
KCl solutions (a) at different pH, (b) at pH 8 containing different concentrations of Ca2+,
(c) at pH 8 containing different concentrations of Mg2+, (d) at pH 11 containing different
concentrations of Ca2+, and (e) at pH 11 containing different concentrations of Mg2+; and
54
(f) Comparison of the Stern potentials of the AFM Si3N4 tip obtained in this this study with
those reported in literature.
In a liquid medium, the charging of a surface mainly occurs in three different ways63,80:
(1) by surface ionization or dissociation (which is the dominant charging mechanism for
oxides, e.g. SiO2, TiO2, Al2O3);
(2) by adsorption or binding of ions from solution (such as ion exchange);
(3) by crystal lattice defects (e.g. isomorphic substitution for clay minerals).
When two surfaces approach each other, if the surface charge comes from irreversible ion
adsorption, dissociation of strong surface groups, or lattice defects, the surface charge
density, which is independent of the surface potential and finally of the separation distance,
remains constant during the approach; if the surface charging mechanism is dominated by
reversible ion adsorption, the surface potential can be assumed constant because the
adsorption-desorption equilibrium responds immediately to every configuration of the
interacting surfaces87. However, constant charge or constant potential are ideal and extreme
cases, and the interaction force between two interacting surfaces in reality would lie in
between.
The charging mechanism for clean silica wafer (which carries many hydroxyl groups in
aqueous solutions) in a solution is chemical adsorption of ions88 and the surface takes up
or releases a proton according to the solution pH:
𝑆𝑖𝑂𝐻2+ 𝑆𝑖𝑂𝐻 + 𝐻+ (takes place at extremely low pH < 3)
55
𝑆𝑖𝑂𝐻 𝑆𝑖𝑂− + 𝐻+
and
𝑆𝑖𝑂− + 𝑀+ 𝑆𝑖𝑂𝑀 (can be adapted for multivalent cations being added into the solution)
Therefore, the silica wafer in inert electrolyte solutions (KCl solution in this case) is
considered to be a constant potential surface. Because silicon nitride is prone to oxidize34
towards SiO2 and the change of its surface charge according to solution pH is due to the
oxidization at its surface81, it is reasonable to assume the surface potential of AFM Si3N4
tip is constant. Consequently, in Figure 4.3, DLVO force curves calculated by constant
surface potential equation were used to fit the experimental data. The determined Stern
potentials of Si3N4 tip are reported in Figure 4.3 (f), showing good agreement with
literature values49,85,86,89.
When Ca2+ and Mg2+ are added into the 10 mM KCl solution, the adsorption mechanisms
of metal cations on oxide/clay surface in aqueous solutions mainly consist of: electrostatic
attraction and surface complexation or specific adsorption at the solid-solution interface.
The adsorption of metal ions at the solid-solution interface is highly pH-dependent. In pH
8 KCl solutions, there is no formation of Ca(OH)+ and very few Mg(OH)+, so there is
almost no specific adsorption of metal cations on the solid surface63,90. However, the
amphoteric SOH surface groups (where S refers to surface Si or Al atoms) on oxide or clay
surfaces can form outer-sphere surface complexes with metal cations91, which can be
expressed with the following equation: ≡ SOH + M(H2O)62+
↔ SO(H2O)M(H2O)5+
+
H+, (where M represents Ca or Mg). The strong surface complexes had a significant effect
on the Stern potential of oxide/clay surface, which also impacted the selection of the
56
equation used in the fitting process. A clean silica wafer, which has many hydroxyl groups
on its surface, is greatly affected by metal cations. As a consequence, the interaction force
curves between the silica wafer and the AFM tip were better fitted with the mixed BC
equation rather than the constant surface potential equation when divalent cations were
added into the KCl solution.
At pH 11, there were more and more Ca(OH)+ and Mg(OH)+ as higher concentration of
Ca2+ and Mg2+ were added to the KCl solutions. According to the speciation diagram (see
Figure C1 in Appendix C), Ca(OH)+ starts to be present at about pH 10.3 and keeps
increasing as solution pH increases to 11, however, the first-order metal hydroxyl species
of magnesium starts to exist at near pH 6 and reaches its peak at pH 10, and then starts to
form metal hydroxide precipitates90. The magnesium hydroxide precipitates at pH 11
would disturb the laser deflection signal, hindering the force measurements when 5 mM
Mg2+ was added into the background electrolyte at pH 11. As more and more irreversible
specific adsorption of first-order metal hydroxyl species occurred on the silica wafer, the
data became better described by the constant surface charge density equation instead of the
mixed BC equation.
57
Figure 4.4 (a) Representative AFM image of a silica wafer surface; and typical interaction
forces (solid lines represent theoretical DLVO forces, open symbols represent
experimental data) between the AFM Si tip and the silica wafer in 10 mM KCl solutions
(b) at different pH, (c) at pH 8 containing different concentrations of Ca2+, (d) at pH 8
58
containing different concentrations of Mg2+, (e) at pH 11 containing different
concentrations of Ca2+, and (f) at pH 11 containing different concentrations of Mg2+.
Figure 4.4 (a) shows a representative image of a silica wafer surface used for colloidal
force measurements, and the root-mean square roughness (Rq) of the silica wafer was 0.22
nm over a 25.0 μm2 area. The results in Figure 4.4 (b) showed that the AFM Si tip was
positively charged at pH 3, and negatively charged at pH 5-10, which means that the PZC
of silicon tip was about pH 3 in 10 mM KCl solutions. This conclusion is consistent with
reported PZC of silicon powder, which was pH 2.892 in 10 mM KNO3 solutions.
For 10 mM KCl solutions, experimental force curves were found to be better fitted by the
mixed BC equation than the constant surface potential equation. The derived Stern
potentials of silicon tip were also different from those of silica wafer. These two findings
indicated that silicon did not carry a constant surface potential like silica, which is probably
because silicon was not fully oxidized to SiO2 or due to the semiconductor nature of silicon.
Under the effect of Ca2+ and Mg2+, the equation adopted in the fitting process gradually
changed from mixed BC equation to constant surface charge density equation. The change
of fitting equation is based on the corresponding charging mechanisms of the tip and the
silica wafer. At pH 8, Ca2+ and Mg2+ would form surface complex with the Si-OH groups
on tip and silica wafer surfaces. However, at pH 11, the first-order metal hydroxyl species
of calcium and magnesium ions specifically adsorbed on the Si-OH groups. When the
concentration of calcium and magnesium ions were increased to be higher than 0.1 mM,
more specific adsorption happened, and caused both tip and silica wafer surfaces to carry
constant surface charges.
59
4.2.2 Characterization of kaolinite samples
4.2.2.1 Samples of kaolinite basal planes
Figure 4.5 SEM images of kaolinite particles deposited on (a) glass substrate and (b)
sapphire substrate after drying on a hot plate; AFM images of kaolinite particles deposited
on (c) glass substrate and (d) sapphire substrate, after the substrate was rinsed with Milli-
Q water and blown dry.
To obtain accurate and reproducible AFM force curves, it is vital to know that the
measurements are performed on the right basal surface of kaolinite particle, and that the
surfaces are smooth enough (surface roughness less than 1 nm). SEM images of the
substrates with deposited kaolinite particles show that the kaolinite basal planes are very
60
flat (see images (a) and (b) in Figure 4.5), so they can orderly stack layer by layer, with a
negatively charged kaolinite Si-basal plane of one kaolinite particle attached to the
positively charged kaolinite Al-basal plane of another kaolinite particle. Finally, the
positively charged Al-basal plane of the last layer of kaolinite particles will attach to the
negatively charged glass substrate, so that the top layer surfaces would be the Si-basal
planes of kaolinite particles, vice versa, when kaolinite particles are deposited on a
positively charged sapphire substrate, the Al-basal planes of kaolinite particles are exposed
at the top layer. It can be seen from the images that the exposed kaolinite basal planes are
flat enough for AFM force measurement, which are confirmed by the AFM images shown
in Figures 4.5 (c) and 4.5 (d). The root-mean-square roughness (Rq) of selected locations
on both imaged kaolinite particles deposited on glass substrate and sapphire substrate was
determined by AFM to be lower than 1 nm.
Because super-fine particles are loosely attached to the substrate, they need to be rinsed off
by large amounts of Milli-Q water before the AFM force measurements. It can be seen
from Figures 4.5 (c) and 4.5 (d) that most of the particles that can stay on the substrate after
rinsing, and especially after scanning by AFM tip, are 1-2 µm in size because they have a
larger contact area with the substrate, which can provide more electrostatic attraction
between the kaolinite particles and the substrate. Micron size particles are also better than
nano size particles for AFM force measurement, because the interaction force between the
AFM tip and the kaolinite surfaces would be less affected by the substrate.
To make sure the opposite basal surfaces of kaolinite were exposed on glass substrate and
sapphire substrate, respectively, interaction forces were measured between the AFM tip
61
and the particles sitting on the two substrates in 10 mM KCl solution of pH 5, as shown in
Figure 4.6.
0 10 20 30
-0.2
0.0
0.2
0.4
0.6
0.8
1.0
AFM tip
vs
Silica wafer
(Electrolyte: 10 mM KCl, pH 5)
Forc
e (
nN
)
Separation (nm)
(a)
0 5 10 15 20 25 30
-0.2
0.0
0.2
0.4
0.6
0.8
1.0
AFM tip
vs
Particles placed on a glass substrate
(Electrolyte: 10 mM KCl, pH 5)
(b)
Forc
e (
nN
)
Separation (nm) 0 5 10 15 20 25 30
0.0
0.5
1.0
AFM tip
vs
Particles placed on a sapphire substrate
(Electrolyte: 10 mM KCl, pH 5)
(c)
Forc
e (
nN
)
Separation (nm)
Figure 4.6 Interaction forces between an AFM tip and silica wafer (a), kaolinite particles
sitting on a glass substrate (b), and particles sitting on a sapphire substrate (c) in 10 mM
KCl solutions of pH 5 (open symbols represent experimental data, and solid lines represent
theoretical DLVO forces).
Using a SurPASS Electrokinetic Analyzer (Anton Paar USA Inc.), zeta potential of silica
wafer in 10 mM KCl solution of pH 5 was determined to be -32 mV. The repulsive
interaction forces between silica wafer and the AFM tip shown in Figure 4.6 (a) mean that
the AFM tip is negatively charged. As repulsive forces were obtained between the
negatively charged tip and the kaolinite particles attaching to the glass substrate, the
negatively charged Si-basal plane surfaces were obtained. In contrast, the attractive forces
shown in Figure 4.6 (c) indicate a positively charged Al-basal plane on the other side of
kaolinite particles was exposed on the sapphire substrate as anticipated.
62
4.2.2.2 Sample of kaolinite edge surfaces
Figure 4.7 SEM images of (a) kaolinite particles deposited on the resin surface, and (b)
kaolinite edge surfaces prepared by ultramicrotome cutting technique; and (c) AFM image
of kaolinite edge surfaces.
The SEM imaging of kaolinite particles deposited on the resin surface (see Figure 4.7 (a))
shows that most of the particles lie flatly and orderly on the resin surface with kaolinite
basal planes facing up. Therefore, mainly the kaolinite edge surfaces will be exposed after
the resin block sandwiched with kaolinite particles in the middle being cut by
63
ultramicrotome. The aspect ratio of the kaolinite surfaces seen in Figure 4.7 (b) is much
higher than the aspect ratio of kaolinite surfaces seen in Figure 4.7 (a), which proves that
the edge surfaces of kaolinite particles have been successfully prepared.
Surface roughness is a vital parameter for acquiring accurate and reproducible data from
AFM force measurements. After being cut by ultramicrotome, the surface of the resin block
sandwiched with kaolinite particles in the middle was imaged by AFM and shown in Figure
4.7 (c). The Rq of the resin surface after cutting is less than 1 nm. The whole image Rq of
Figure 4.7 (c) is 18.3 nm over 25 μm2, and the Rq values of local parts vary between 1 to
2 nm, which means the edge surface is smooth enough for nanoscale study on the charging
properties of kaolinite edge surfaces.
4.2.3 Effect of divalent cations on Stern potential of kaolinite Si-basal
planes
AFM images were taken before and after the force measurements to make sure the kaolinite
particles were still attached to the substrate. If the particles were no longer attached onto
the substrate after the force measurements, the force curves taken on those particles would
not be used in the following fitting process.
64
Figure 4.8 Interaction forces between the AFM Si3N4 tip and the kaolinite Si-basal plane
in 10 mM KCl solutions (a) at different pH, (b) at pH 8 containing different concentrations
of Ca2+, (c) at pH 8 containing different concentrations of Mg2+, (d) at pH 11 containing
different concentrations of Ca2+, and (e) at pH 11 containing different concentrations of
65
Mg2+; and (f) Summary of the Stern potential of kaolinite Si-basal plane at pH 8 and 11 in
10 mM KCl solutions containing different concentrations of Ca2+ and Mg2+.
Typical approaching force curves between AFM tip and kaolinite Si-basal plane in 10 mM
KCl solution under various pH environments are shown in Figure 4.8 (a). The siloxane
surface of kaolinite carries constant surface charge20 because: 1) the surface charge of the
siloxane surface mainly arises from isomorphic substitution, which leads to a permanent-
negative charge; and 2) the hydrolysis of siloxane bonds is slow unless at extremely high
pH. Therefore, a mix BC equation was applied to fit the experimental forces between the
AFM Si3N4 tip with a constant surface potential and the kaolinite siloxane basal plane with
a constant surface charge. For kaolinite Si-basal plane, repulsive interactions were found
from pH 5 to pH 10, and the attractive interaction force occurred at pH 3 was attributed to
the positively charged AFM Si3N4 tip under this condition. Therefore, kaolinite Si-basal
plane carried negative charges from pH 3 to pH 10 in 10 mM KCl solutions. Moreover,
because the Stern potentials of Si-basal plane changed dramatically at pH 3, the PZC of the
siloxane surface of kaolinite was determined to be below, and near pH 3. According to the
results in this study, the Stern potential of kaolinite siloxane surface almost did not change
except at pH 3, which proves that the Si-basal plane of kaolinite carries a permanent surface
charge that is negative and pH-insensitive20.
At pH 8, 10 mM KCl solutions with various concentrations of divalent cations (0, 0.1, 0.5,
1, and 5 mM) were used as the background electrolytes. The interaction forces between
AFM tip and kaolinite Si-basal plane were measured and shown in Figures 4.8 (b) and (c).
The long-range repulsion was found in all conditions but was depressed with the increasing
addition of divalent cations. When the concentration of Ca2+ or Mg2+ was increased to 5
66
mM, the interaction force was almost negligible, yet still repulsive. According to the Stern
potentials, kaolinite Si-basal plane was found to be relatively less affected by the divalent
cations compared to the silica wafer because the siloxane groups on this basal plane are
hard to be hydrolyzed into ≡ SiOH, thus they are harder to form surface complexes with
metal cations. Hence, the decrease of the Stern potentials of kaolinite Si-basal plane with
the increase of divalent cation concentrations is attributed to the increasing electrostatic
interaction between the negatively charged basal plane and the divalent cations. When the
concentration of metal cations was increased to 5 mM, the local concentration of metal
cations on the surface became as high as [M2+]0 ≈ 5 × 10−3𝑒+200∕25.7 ≈ 12M80. At such
high surface concentration, divalent ions can chemically bind to negative surface sites80,
therefore considerably lowering the Stern potential of kaolinite Si-basal plane. Comparing
Figures 4.8 (b) and (c), Ca2+ and Mg2+ have similar magnitude of effect on kaolinite Si-
basal plane.
At pH 11, when specific adsorption of Ca(OH)+ and Mg(OH)+ on kaolinite Si-basal plane
becomes more and more significant as the increase of initial divalent cation concentration,
the experimental force curves should be fitted by the theoretical DLVO force curves
calculated under constant charge density. Charge reversal was found at 1 mM Mg2+
addition, but not with any concentrations of Ca2+, indicating a stronger impact of Mg2+ than
Ca2+ at pH 11. This phenomenon was also observed by Gan et al.90 on their investigation
of the impact of divalent cations on the zeta potential of quartz. In the sulfonate flotation,
at pH 11, the flotation recovery of sulfonate in the presence of 0.1 mM magnesium ions
could reach 100%, while the flotation recovery was only about 20% with the same
concentration of calcium ions31.
67
According to the speciation diagram for 1 mM Ca2+ and Mg2+ (see Figure C1 in Appendix
C), the concentration of Ca(OH)+ and Mg(OH)+ are both about 10-5 mol/L at pH 11.
However, the hydroxide precipitate amounts of Ca and Mg vary a lot, which are 10-8 mol/L
and 10-5 mol/L, respectively. Therefore, the difference between the impact of Ca2+ and
Mg2+ at pH 11 is probably caused by the magnesium hydroxide precipitate on kaolinite Si-
basal plane. For the Stern potential of kaolinite basal plane to reverse sign by Mg2+ addition,
the suggested mechanism involves nucleation and growth of magnesium hydroxide
precipitate at the basal surfaces. Following the formation of the precipitate, the basal
surfaces of kaolinite are postulated to have the electrokinetic features of magnesium
hydroxide93, and this would result in more charge in the Stern layer94.
Mg(OH)2 was found to strongly adsorb on the quartz surfaces by Gan et al.90, because their
XPS results of Mg adsorbed on the glass slide at pH 10.9 before and after rinsing showed
that the surface atomic concentration of Mg still remained high after rinsing.
68
4.2.4 Effect of divalent cations on Stern potential of kaolinite Al-basal
planes
Figure 4.9 Interaction forces between the AFM Si3N4 tip and the Al-basal plane of kaolinite
in 10 mM KCl solutions (a) at different pH, (b) at pH 8 containing different concentrations
of Ca2+, (c) at pH 8 containing different concentrations of Mg2+, (d) at pH 11 containing
69
different concentrations of Ca2+, and (e) at pH 11 containing different concentrations of
Mg2+; and (f) Summary of the Stern potential of kaolinite Al-basal plane at pH 8 and 11 in
10 mM KCl solutions containing different concentrations of Ca2+ and Mg2+.
Interaction force curves between the AFM tip and the aluminum oxy-hydroxyl surface of
kaolinite (Al-O-OH) in 10 mM KCl solution under various pH environments are shown in
Figure 4.9 (a). The protonation and deprotonation of the exposed Al-OH surface groups
are responsible for the formation of a net surface charge. However, constant surface
potential equation could hardly be used to fit the experimental force curves. A better fitting
result was obtained using the mixed BC equation, which may be caused by: 1) isomorphic
substitution of Al3+ by Mg2+ in the octahedron sheet; 2) the pulling effect of surface groups
from the T sheet. Repulsive interaction force found at pH 3 indicates that Al-basal plane of
kaolinite is positively charged under this condition. Attraction force curves occurred at pH
5 and 6 meaning that this surface still carries positive charge. However, when the solution
pH was increased to 8 and 10, the kaolinite Al-basal plane became negatively charged
because repulsive force curves were obtained between the negatively charged AFM tip and
the particle surface. Hence, the PZC of kaolinite Al-basal plane was found to be between
pH 6 and 8, which agrees well with the result reported by Gupta et al35. These findings also
proved that the Al-basal plane is slightly pH-sensitive because its Stern potential changed
only slightly, except when the solution pH was close to its PZC.
Significant impact of divalent cations on the amphoteric Al-OH surface groups was
observed due to the formation of surface complexes (AlO(H2O)Ca(H2O)5+ and
AlO(H2O)Mg(H2O)5+) and electrostatic attraction. The mixed BC equation was adopted in
the fitting process. Figure 4.9 (b) reveals the influence of Ca2+ on the charging
70
characteristics of kaolinite Al-basal plane. When the concentration of Ca2+ was increased
from 0 to 0.1 mM, and to 0.5 mM, the interaction forces were repulsive, but when the
addition of Ca2+ was further increased to 1 mM, the interaction forces became net attractive,
which demonstrates that the Stern potential of kaolinite Al-basal plane was reversed from
negative to positive at the addition of 1 mM Ca2+ in 10 mM KCl solutions at pH 8. The
interaction force reduced to almost zero when the addition of Ca2+ was increased to 5 mM
due to large extent of the electrical double-layer compression. However, the interaction
forces between the AFM tip and the kaolinite Al-basal plane were found to be
monotonically repulsive with the addition of Mg2+ varying from 0.1 to 5 mM at pH 8. The
addition of 1mM Ca2+ resulted in charge reversal of the Al-basal plane while 1 mM Mg2+
did not. This implies that Ca2+ has a stronger effect than Mg2+ on the Stern potential of
kaolinite Al-basal plane at pH 8. According to the quantitative model for the adsorption of
hydrolysable metal ions on oxide-liquid interface proposed by James and Healy69, the free
energy change of adsorption is the sum of the change in coulombic energy, the change in
secondary solvation energy and the change in a specific ‘chemical’ interaction, that is:
∆𝐺0𝑎𝑑𝑠𝑖
= ∆𝐺0𝑐𝑜𝑢𝑙𝑖
+ ∆𝐺0𝑠𝑜𝑙𝑣𝑖
+ ∆𝐺0𝑐ℎ𝑒𝑚𝑖
. At pH 8, ∆𝐺0𝑐ℎ𝑒𝑚𝑖
is negligible because
there is almost no specific adsorption reaction for both cations. Generally, for two metal
ion species of the same charge, the larger the ionic dimension, the more negative the sum
of the coulombic and solvation energy change, which means larger ion is more favorable
to the adsorption process. Moreover, Yan et al.85 reported that the less hydrated Ca2+ could
be more proximal to the oxide surface than Mg2+ due to its lower hydration enthalpy. As a
result, compared to Mg2+, Ca2+, which has a larger ionic dimension and is less hydrated,
71
gets closer to the clay surface, thus generating a stronger impact on the Stern potential of
kaolinite Al-basal plane.
Comparing (b) and (d), (c) and (e) in Figure 4.9, the effect of the first-order metal hydroxyl
groups on the Al-basal surface is clearly seen when the solution pH was increased from 8
to 11. The surface charge reversal point for Ca2+ addition are the same (1mM addition) at
pH 8 and 11, but because the concentration of Ca(OH)+ increases significantly with
solution pH, the main binding mechanism with Al-OH changes from surface complexion
of Ca(H2O)62+ to specific adsorption of Ca(OH)+. Therefore, the equation needed to fit the
experimental forces changes from mixed BC equation at pH 8 to constant surface charge
density equation at pH 11. Charge reversal occurred at 0.5 mM Mg2+ addition at pH 11
implies that Mg(OH)+ becomes dominant, which leads to the formation of a strong covalent
bond (AlOMg+) with the surface Al-OH groups. Magnesium ions were also found to have
bigger impact than calcium ions on the Stern potential of kaolinite Al-basal plane at pH 11,
probably due to the formation of magnesium hydroxide precipitate on the surface.
72
4.2.5 Effect of divalent cations on Stern potential of kaolinite edge
surfaces
Figure 4.10 Typical interaction forces between the AFM Si tip and kaolinite edge surface
in10 mM KCl solutions (a) at different pH, (b) at pH 8 containing different concentrations
73
of Ca2+, (c) at pH 8 containing different concentrations of Mg2+, (d) at pH 11 containing
different concentrations of Ca2+, and (e) at pH 11 containing different concentrations of
Mg2+; and (f) summary of the Stern potentials of kaolinite edge surfaces.
Interaction force curves between AFM Si tip and the edge surfaces of kaolinite shown in
Figure 4.10 (a) manifest that the edge surface of kaolinite was positively charged at pH 3,
and negatively charged from pH 6 to 10 in 10 mM KCl solutions. The PZC of the kaolinite
edge surface was found to be around pH 5 because positive, zero, and negative surface
potentials were all obtained during the fitting process. Because the surface groups of
kaolinite edge surface are broken Si-O and Al-O bonds, which are easily hydrolyzed into
Si-OH and Al-OH in KCl solutions. Si-OH and Al-OH would be protonated or
deprotonated according to solution pH, so kaolinite edge surface carries a constant surface
potential. Therefore, mixed BC equation was used to fit the experimental force curves in
10 mM KCl solutions.
In this research, the impacts of Ca2+ and Mg2+ on the surface potential of kaolinite edge
surface were explored for the first time. Interaction force curves between the edge surfaces
of kaolinite and the AFM Si tip in 10 mM KCl solutions with varying concentration of
divalent ions are shown in Figure 4.10 (b)-(e). The interaction force curves were repulsive
at the additions of 0 and 0.1 mM Ca2+, which became attractive when the concentration of
Ca2+ was increased to 0.5, 1 mM and 5 mM at pH 8. The interaction forces were still
repulsive at 0.5 mM Mg2+ addition, albeit very weak. This means the impact of Mg2+ is
smaller compared to Ca2+, which matches well with what we found for kaolinite basal
planes. At pH 11, when the concentration of divalent ions was increased to 0.5 mM, the
fitting equation was changed from mixed boundary equation to constant surface charge
74
density equation because more specific adsorption happened. Ca2+ can reverse the surface
charge sign at 0.5 mM addition, pH 11, due to specific adsorption. Again, the Stern
potential of about 20 mM was obtained with 1 mM Mg2+ addition at pH 11, which is caused
by magnesium hydroxide precipitation.
What needs to be mentioned is that typical force curves between AFM Si tip and kaolinite
edge surfaces are quite different from the force curves between AFM Si tip and resin
surface/cavity, which are extremely weak and very bumpy, respectively.
4.3 Summary
Following the successful preparation of micron-sized and smooth kaolinite basal planes,
the Stern potentials of different surfaces were studied in 10 mM KCl solutions in the
absence and presence of divalent cations (Ca2+ and Mg2+, respectively) at different pH. The
Stern potentials were derived by fitting the calculated DLVO force curves with the
measured interaction force curves between the AFM tip and the basal planes. Proper
boundary conditions were applied during the fitting processes. According to the
experimental results, we can achieve the following understanding of the anisotropic surface
charging properties and mechanisms of different kaolinite surfaces:
(1) The Stern potentials of the siloxane basal plane of kaolinite were negative from pH 3
to 10 in 10 mM KCl solutions, which barely changed with the variation of solution pH.
The aluminum oxy-hydroxyl basal plane of kaolinite was slightly pH-sensitive and its PZC
was found to be between pH 6 to 8. The pH-sensitive kaolinite edge surfaces had a PZC of
about pH 5. These findings match well with previous research and generally accepted
theory35,36.
75
(2) For the siloxane basal plane of kaolinite, when Ca2+ was added into the 10 mM KCl
solution at pH 8 and 11, the Stern potential became less negative without changing to
positive with the additions increasing from 0.1 to 5 mM. When varying concentrations of
Mg2+ were added into the background electrolyte, although the Stern potential remained
negative at pH 8, it became positive at 1 mM Mg2+ addition at pH 11.
(3) For the aluminum oxy-hydroxyl basal plane of kaolinite, when the concentration of
Ca2+ was increased from 0.1 to 0.5 mM at pH 8 and 11, the Stern potentials were both
depressed without changing from negative to positive. When the addition of Ca2+ was
further increased to 1 mM, the Stern potential of Al-basal plane was reversed to positive at
both pHs. In the study of Mg2+, the addition of Mg2+ did not have the ability to achieve
charge reversal at pH 8, however, Mg2+ began to make serious impact on the Stern potential
of the Al-basal plane at pH 11. At pH 11, 0.5 mM addition of Mg2+ could reverse the Stern
potential from negative to positive.
(4) For kaolinite edge surfaces, the addition of Ca2+ can cause charge reversal at 1 mM and
0.5 mM, at pH 8 and 11, respectively. Charge reversal did not happen with the addition of
Mg2+ at pH 8. However, with more formation of Mg(OH)+ at pH 11, stronger specific
adsorption reversed the surface charge sign of kaolinite edge at 1 mM addition.
(5) The Stern potentials of three kaolinite basal planes responded differently to solution pH
due to different dominant charging mechanisms: isomorphic substitution for the Si-basal
plane and protonation/deprotonation for the Al-basal plane and edge surfaces. At pH 8, the
distinct responses to divalent cations of the three surfaces imply distinct binding
mechanisms: electrostatic attraction dominates for the Si-basal plane and the formation of
76
strong surface complexes prevails on the Al-basal plane and edge surfaces (reactions and
schematic of the adsorption mechanisms at pH 8 can be seen in Figure 4.11). Moreover,
Ca2+ was found to be more favorable to the adsorption process than Mg2+ at pH 8, which
is attributed to the larger ionic dimension and less hydrated nature of calcium ions than
magnesium ions. At pH 11, Mg2+ had more significant impact on both basal planes than
Ca2+, and this is due to the nucleation and growth of magnesium hydroxide precipitate at
the basal surfaces (reactions and schematic of the adsorption mechanisms at pH 11 can be
seen in Figure 4.12). With the formation of the precipitate, the kaolinite basal and edge
surfaces start to have the electrokinetic features of magnesium hydroxide, which accounts
for the charge reversal occurred on all three kinds of surfaces with 1 mM Mg2+ addition at
pH 11.
The possible reactions between the studied ions with kaolinite surfaces under different
liquid environments were summarized as follows, and the corresponding schematics of the
ion adsorption on kaolinite surfaces were illustrated in Figure 4.11 and 4.12.
At pH 8:
Electrostatic attraction: ∥ XO− + M(H2O)62+
↔∥ XO−M(H2O)62+
(4.1)
Outer-sphere complex: ∥ XOH + M(H2O)62+ →∥ XO(H2O)M(H2O)5
+ + H+ (4.2)
Here, X stands for Si or Al.
77
Figure 4.11 Schematics of the adsorption of divalent cations on kaolinite surfaces at pH 8.
At pH 11:
Specific adsorption: ∥ XOH + M(OH)+ →∥ XOM+ + HOH (4.3)
Surface precipitation: ∥ XOH + Mg2+ + 2H2O →∥ XOH[Mg(OH)2] + 2H+ (4.4)
Figure 4.12 Schematics of the adsorption of divalent cations on kaolinite surfaces at pH
11.
78
In conclusion, we investigated the influence of divalent cations (Ca2+ and Mg2+) on the
anisotropic charging characteristics of kaolinite, which is a valuable clay mineral and has
abundant applications. The new information on the Stern potentials of kaolinite basal
planes in 10 mM KCl solutions with the presence of Ca2+ and Mg2+ gives insights into their
adsorption mechanisms on distinct basal planes. Direct force measurements using AFM on
carefully prepared kaolinite basal planes described in this research can be extended to other
phyllosilicates to explore their surface properties and colloidal behaviors, which could lead
to great contributions to the processing and utilization of broad natural resources.
79
Chapter 5 Imaging of Ion Adsorptions on Kaolinite Basal
Planes
5.1 Introduction
In this chapter, atomic resolution imaging was performed on kaolinite basal planes to
directly see the ion distributions on kaolinite surfaces. By this model-independent way,
proposed ion adsorption mechanisms (in Chapter 4) were confirmed.
5.2 Results and discussion
In chapter 4, we reported the effect of divalent cations on the Stern potentials of kaolinite
surfaces, and brought up possible ion adsorption mechanisms under different solution
conditions. To further confirm our speculation, atomic resolution imaging was performed
on kaolinite basal surfaces to explore the ion distribution. The atomic resolution imaging
technique was described in Chapter 3, and the imaging processing procedure was provided
in Appendix A.
80
5.2.1 Ion adsorption of monovalent ions
Figure 5.1 Surface lattice structure of kaolinite: Top view of the T sheet (a), and atomic
resolution images of kaolinite Si-basal plane in Milli-Q water (b) and 10 mM KCl solution
(c) at pH 5; Top view of the O sheet (d), and atomic resolution images of kaolinite Al-basal
plane in Milli-Q water (e) and 10 mM KCl solution (f) at pH 5; with insets being high
resolution views (top) and 2D spectrum (Fast Fourier Transform) image of the same data
(bottom) (large white, solid circle represents oxygen; large, dotted circle shows the fourth
oxygen atom of each tetrahedron pointing downward; small red circle represents silicon;
large grey, solid circle represents hydroxyl; small blue circle represents aluminum).
The top view of the T sheet (Figure 5.1 (a)) and O sheet (Figure 5.1 (d)) shows hexagonal
rings composed of outer oxygen atoms, with silicon/aluminum atoms residing below the
81
top oxygen plane. The hexagonal rings were clearly observed from the atomic resolution
images of both kaolinite Si-basal plane and kaolinite Al-basal plane taken in ultrapure
Milli-Q water, as shown in Figures 5.1 (b) and (e). In Figure 5.1 (b), the spacing between
unit cells in x and y directions for kaolinite Si-basal plane was 0.514 nm and 0.996 nm,
respectively. In Figure 5.1 (e), the spacing between unit cells for kaolinite Al-basal plane
was 0.481 nm and 0.939 nm in x and y directions, respectively (the measurements of the
spacing are provided in the supporting information). These spacing values agree well with
the values reported in literature, which are given in Table 5.1.
Table 5.1 Comparison of the unit cell spacing of kaolinite basal planes
Method Kaolinite Si-basal plane Kaolinite Al-basal plane Reference
XRD x = 0.515 nm y = 0.89 nm 26
AFM
x = 0.514 nm
y = 0.996 nm
x = 0.481 nm
y = 0.939 nm This work
x = 0.51 nm
y = 0.92 nm
x = 0.50 nm
y = 0.905 nm
74
x = 0.50 ± 0.04 nm x = 0.36 ± 0.04 nm 95
The adsorption of K+ and Cl- on oppositely charged kaolinite basal planes was studied at
pH 5. According to Hunter63, specifically adsorbed ions can penetrate into the inner region
to have a strong and short-range interaction with the surface. If K+ or Cl- can adsorb
specifically at the interface, the strong binding between K+ and Si-basal plane or between
Cl- and Al-basal plane would be able to alter the surface lattice structure. However,
comparison of Figures 5.1 (b) and (c), (e) and (f) shows no visible change in the surface
82
lattice structure. Therefore, strong binding of monovalent ions (K+ and Cl-) on negatively
and positively charged clay surfaces were not detected, regardless whether they were
hydrated or not.
5.2.2 Ion adsorption of divalent ions
In Chapter 4, kaolinite Si-basal plane and Al-basal plane were determined to be negatively
charged in KCl solutions at pH above 8. According to the speciation diagram (see Figure
C1 in Appendix C) of 1 mM Ca2+, the concentration of Ca(OH)+ is negligible at pH < 8,
and it increases dramatically with solution pH. The surfaces of the two basal planes were
imaged when 1 mM Ca2+ was added into the KCl solution at pH 8 and 11, respectively.
Figures 5.2 (a) and (b) show the surface structures of the two basal planes in 10 mM KCl
+ 1 mM Ca2+ aqueous solutions of pH 8. The periodic hexagonal rings remained on Si-
basal planes as shown in Figure 5.2 (a) although it was not as sharp as in the case without
calcium ions as shown in Figure 5.1 (c). In contrast, a distorted structure was seen on the
Al-basal plane as shown in Figure 5.2 (b) in comparison with the case in the absence of
calcium as shown in Figure 5.1 (e). However, the hexagonal rings completely disappeared
on the Si-basal plane, so did the distortion on Al-basal plane, which were replaced by clear,
periodic and rectangular structures on both surfaces when the solution pH was increased to
11.
83
Figure 5.2 Atomic resolution images of kaolinite Si-basal (a), and Al-basal plane (b) in 10
mM KCl + 1 mM Ca2+ at pH 8; kaolinite Si-basal plane (c), and Al-basal plane (d) in 10
mM KCl + 1 mM Ca2+ at pH 11.
Although both basal planes were negatively charged at pH 8, the surface charging
mechanisms are completely different. The surface siloxane (SiOSi) groups of Si-basal
plane could be considered inert with the negative surface charges permanently carried by
the Si-basal plane being mainly from isomorphic substitution. When Ca2+ were added at
pH 8, the highly hydrated divalent calcium ions (Ca(H2O)62+) could only interact with the
negatively charged surface through electrostatic attraction in the diffuse electrical double
layer. When the AFM tip was scanning over the surface, the tip would easily push the
84
hydrated calcium ions away, which explains why the hexagonal surface structure was not
affected. On the contrary, the amphoteric Al-OH surface groups on the Al-basal plane are
pH-responsive and can form surface complexes with hydrated calcium ions91. The outer-
sphere complexes (AlO(H2O)Ca(H2O)5+) would interfere with the hexagonal structure seen
on bare Al-basal plane as revealed in Figure 5.2 (b). The scanning of the AFM tip over the
charged solid surface is schematically illustrated in Figure 5.3.
Figure 5.3 Schematic of AFM tip scanning surface in solution.
When the solution pH was 11, both basal planes showed rectangular structures. Because
calcium ions would be highly hydrated in solution, they are unable to penetrate into the
inner region and interact directly with the surfaces63. Therefore, the structure change
coming from very strong surface binding should be attributed to the Ca(OH)+ formed at
high pH. The strong binding of Ca(OH)+ at pH 11.7 was also seen on a glass slide by Gan
85
et al.90 using XPS, which showed a rather higher ratio of Ca/Si after the glass slide was
blown dry compared to the ratio obtained at pH 6. Therefore, it is only possible that the
strong binding of Ca(OH)+ (that can resist air blowing or AFM scanning) caused the change
of the surface structure. This strong, direct surface binding of Ca(OH)+ on kaolinite basal
planes was the so-called specific adsorption.
5.3 Summary
According to the atomic resolution imaging on kaolinite surfaces under different
background electrolytes, monovalent ions were not able to alter the surface lattice structure
of kaolinite basal surfaces, no matter whether they were hydrated or not. The observed
change in morphology of kaolinite basal plane surfaces in calcium solutions of increasing
pH suggests strong and chemical binding of divalent cations on clay surfaces, mostly in
the form of surface complexation and specific adsorption under low pH and high pH
solution environment, respectively, following possible reactions (4.1), (4.2), and (4.3)
proposed in Chapter 4.
86
Chapter 6 Understanding Rheology and Settling Behaviors of
Kaolinite Suspension
6.1 Understanding the rheology behavior of kaolinite suspension
6.1.1 Calculation of interaction energy
After determining the Stern potentials of distinct kaolinite surfaces, the interaction energy
between different surfaces could be calculated based on DLVO theory.
The van der Waals interaction energy per unit area between planar surfaces is calculated
by:
𝑈𝑣𝑑𝑤 = −𝐴
12𝜋𝐷2 (6.1)
The electrical double-layer interaction energy per unit area between two planar surfaces is
given by the following equations85,96,97 under different boundary conditions:
For constant surface potential BC:
𝑈𝜓−𝜓 =𝜀𝜀0𝜅
2(𝜓𝑎
2 + 𝜓𝑏2)[1 − 𝑐𝑜𝑡ℎ(𝜅𝐷)] + 2𝜓𝑎𝜓𝑏𝑐𝑜𝑠𝑒𝑐ℎ(𝜅𝐷) (6.2)
For constant surface charge density BC:
𝑈𝜎−𝜎 =1
2𝜀𝜀0𝜅(𝜎𝑎
2 + 𝜎𝑏2)[𝑐𝑜𝑡ℎ(𝜅𝐷) − 1] + 2𝜎𝑎𝜎𝑏𝑐𝑜𝑠𝑒𝑐ℎ(𝜅𝐷) (6.3)
For mixed BC:
𝑈𝜎−𝜓 =1
22𝜓𝑎𝜎𝑏𝑠𝑒𝑐ℎ(𝜅𝐷) + (
𝜎𝑏2
𝜀𝜀0𝜅− 𝜀𝜀0𝜅𝜓𝑎
2) × [𝑡𝑎𝑛ℎ(𝜅𝐷) − 1] (6.4)
87
6.1.2 Shear yield stress measurement
Kaolinite suspension (35 wt%) was prepared in 10 mM KCl solutions and thoroughly
mixed by IKA RW-20 overhead digital impeller (IKA Works, Inc., US) at 600 rpm for 30
min. The impeller was gently removed from the suspension, and the suspension was slowly
poured into a beaker for pH adjustment. The suspension was put into a vacuum aspirator
for one hour to remove the entrained air, and left overnight before proceeding to rheology
testing.
An AR-G2 rheometer (TA instruments, DE) and a concentric cylinder geometry were used
to study the rheology. The final pH of the kaolinite suspension was measured before the
shear stress measurement. Rheological properties of kaolinite suspension were obtained by
steadily increasing , then decreasing the shear rate (1 to 10 1/s) of the rotating spindle while
measuring the shear stress. The suspension showed non-Newtonian behavior and the yield
stress was obtained by extrapolation from the linear part of the rheometer data, using the
Bingham model98:
τ = 𝜏𝐵 + 𝜂𝑃𝛾 (6.5)
where, τ is the shear stress, 𝜏𝐵 is the Bingham yield stress, 𝜂𝑃 is the plastic viscosity and
𝛾 is the shear rate.
88
6.1.3 Understanding kaolinite suspension rheology behavior through
interaction energy calculation
Figure 6.1 Impact of solution pH on the shear yield stress of kaolinite suspension.
The shear yield stress of kaolinite suspension was investigated in 10 mM KCl solutions
from pH 3 to 10. The shear yield stress was about 13 Pa at pH 3, which increased to a
maximum value of 15.4 Pa at pH 3.6. The shear yield stress started to decrease sharply
from pH 3.6 to pH 4.3, and became almost neglectable from pH 5 to pH 10.
Kaolinite has three different surfaces: Si-basal plane, Al-basal plane and edge surface. In a
kaolinite suspension, different surfaces would interact with each other, thus forming six
kinds of inter-particle pairs: Si-basal plane vs Si-basal plane (Si vs Si), Al-basal plane vs
Al-basal plane (Al vs Al), edge surface vs edge surface (E vs E), Si-basal plane vs Al-basal
plane (Si vs Al), Si-basal plane vs edge surface (Si vs E) and Al-basal plane vs edge surface
(Al vs E). Based on the Stern potential results in 10 mM KCl solutions at different pH, the
0
2
4
6
8
10
12
14
16
18
3 5 7 9 11
Sh
ear
Yie
ld S
tres
s (P
a)
pH
89
interaction energy per unit area for these six kinds of associations was calculated and
shown in Figure 6.2.
Figure 6.2 Interaction energy/unit area between different kaolinite surfaces in 10 mM KCl
solutions at different pHs.
90
The aggregation structure among charged clay particles is mainly determined by the
repulsive and attractive forces between two surfaces. We can see from Figure 6.2 (a) that,
at pH 3, the repulsive force between Si-basal plane and Si-basal plane is significantly
smaller than all other cases. The net repulsive forces between Al-basal planes and edge
surfaces (shown in Figure 6.2 (b) and (c)) at pH 3, 8 and 10 are not very different, and they
gradually change to small attraction at pH 5 and 6. Although the interaction energy was
calculated between two same surfaces and they carried the same surface charges/Stern
potentials, the interaction energy could be attractive when the Stern potentials were low
(pH 5 and 6, in this case). This is because the attractive VDW force overcomes the repulsive
EDL force, and becomes dominant in these cases. In Figure 6.2 (d), we can see that the
attraction between Si-basal plane and the Al-basal plane became significant at pH 3 and 5,
especially at pH 5. Attractive interaction was obtained when Al-basal plane interacted with
edge surface at pH 5 and 6, which became slightly repulsive at pH 3 and 8, and the repulsion
reached its maximum at pH 10. The interaction between Si-basal plane and edge surfaces
gradually changed from attractive to repulsive as the solution pH increased from 3 to 10.
Based on the calculated interaction energy profiles given in Figure 6.2, we can predict the
pH of the maximum shear yield stress for different pairs of surfaces, which is listed in
Table 6.1.
91
Table 6.1 Prediction of the pH of the maximum shear yield stress for different pairs of
surfaces from the calculated interaction energy profiles.
Pairs of surfaces Predicted pH of the maximum shear yield stress
Si vs Si 3
Al vs Al 6
E vs E 5
Si vs Al 5
Si vs E 5
Al vs E 3
It can be seen from Table 6.1 that, among the six kinds of pairs of surfaces, the maximum
yield stress of two kinds (Si vs Si, and Al vs E) would occur at pH 3, and three kinds (E vs
E, Si vs Al, and Si vs E) would occur at pH 5. Therefore, it is reasonable to predict that the
maximum shear yield stress for the whole system may occur between pH 3 and 5. Moreover,
according to the calculation results showed in Figure 6.2, it can be concluded that at pH 3,
the interaction among kaolinite particles in a suspension was dominated by big attraction
between Si-basal plane vs edge surfaces, and Si-basal plane vs Al-basal plane. The
repulsion between same surfaces at pH 3 are relatively small compared to pH 6, 8 and 10,
and it became even smaller when solution pH increased from 3 to 5. The particles tend to
come closer and form card house structure (see Figure 6.3 (a)) via attractions between basal
planes and edge surfaces at pH around 3, which explains why the highest shear-yield stress
was obtained at pH 3.6 (see Figure 6.1).
At pH 5, the attractive interaction between particles come from Al-basal vs Al-basal, edge
vs edge, Si-basal vs Al-basal and Si-basal vs edge. The shear-yield stress became rather
small but still not negligible, which means the newly formed structure at pH 5 was not the
strong card house structure that can introduce high shear-yield stress. Therefore, the
92
dominated attraction should come from Si-basal plane vs Al-basal plane and Al-basal plane
vs Al-basal plane, and the particles would form basal plane vs basal plane (mainly
dominated by Si-basal) stacking structure as illustrated in Figure 6.3 (b).
When the solution pH was further increased to 6, 8 and 10. Net repulsion occurred at almost
all inner particle associations, so the particles could not get closer to each other, and thus
staying dispersed in the solution (see Figure 6.3 (c) for the dispersed structure). This is also
why kaolinite particles are hard to settle down in a pH 8 process water in oil sands tailings.
Figure 6.3 Proposed aggregate structures: (a) card house structure; (b) parallel stacking; (c)
dispersed structure.
6.2 Understanding the settling behavior of kaolinite suspension
6.2.1 Settling test
Because the pH of oil sands tailings is about 7.5 to 8, and at this pH, Ca2+ and Mg2+ have
similar effect on the kaolinite particles, the settling tests were only conducted in 10 mM
KCl solutions with the addition of Ca2+ at pH 8. To investigate the impact of magnesium
precipitate, settling tests were also performed with 5 mM Ca2+ and Mg2+ addition at pH 11.
Kaolinite suspensions (5%) were prepared in 10 mM KCl solutions containing various
concentrations of calcium chloride. The suspensions were mixed by IKA RW-20 overhead
93
digital impeller (IKA Works, Inc., US) at 600 rpm for 5 min. The pH of the well-dispersed
suspensions were adjusted to 8 using 1 M HCl and 1M KOH while being magnetically
stirred. The prepared suspenisons were transferred into graduate cylinders (100 mL) for
settling test.
6.2.2 Understanding kaolinite suspension settling behavior through
interaction energy calculation
The mudline, i.e., the interface between the supernatant and sediments, was recorded as a
function of the settling time. The mudline height was calculated using the absolute height
of the mudline divided by the total height of the suspension. The settling curves describe
the change of the mudline height vs settling time, which were shown in Figure 6.4. The
initial settling rate (ISR) was defined as the initial slope of the settling curve.
Figure 6.4 Settling performance of kaolinite suspension.
94
Figure 6.5 Impact of divalent cations on the interaction energy/unit area between different
kaolinite surfaces.
When Ca2+ were added into the suspension, it can be seen in Figure 6.4 that the IRS
increased very slowly when the concentration of Ca2+ was increased from 0 to 1 mM, but
the IRS increased sharply when the concentration increased to 5 mM. This suggested that
the addition of divalent cations can fasten the settling of kaolinite particles in a kaolinite
suspension when the Stern potentials were highly decreased for all surfaces.
When 0.1 mM Ca2+ were added into the kaolinite suspension, the calculated interaction
energies showed that the repulsion between different surfaces were highly reduced.
However, all of the interaction energies were still repulsive, so the IRS with 0.1 mM
divalent cation addition is almost the same as the IRS in 10 mM KCl solution, and only the
95
final mudline height of 0.1 mM Ca2+ addition is lower than that of 10 mM KCl solution
case (see Figure 6.4).
Because the Stern potentials of kaolinite surfaces were further reduced with 0.5 and 1 mM
Ca2+ addition, the interaction energies between surfaces changed from repulsive to
attractive (see Figure 6.5), even between same surfaces, which is because the attractive
VDW force became dominant. As we can see from Figure 6.4 that, the IRS for 0.5 mM
divalent cation addition was still very close to that of 10 mM KCl solution case, and the
IRS only increased a little at 1 mM divalent cation addition, which was not increased as
expected. This is because the attraction among particles are still dominated by both basal
vs basal and basal vs edge (see Figure 6.5 (b) and (c)), so the structure of the aggregation
formed at these two conditions are card-house structure as shown in Figure 6.6 (b). The
card house structure is able to trap a large amount of water and slows the settling process.
If we keep increasing the addition of divalent cations, the Stern potentials of kaolinite Al-
basal and edge surfaces could be reversed from negative to positive, and the Stern potential
of kaolinite Si-basal planes were also highly depressed according to the findings in Chapter
4. The interaction energies between kaolinite surfaces (see Figure 6.5 (d)) showed strong
attraction except Si-Si, and the attraction among particles is dominated by Al-E and Si-E.
The dominating attraction coming from basal vs edge may cause the formation of a brush-
like aggregation structure, as shown in Figure 6.6 (c). This dense compact will ultimately
lead to a fast settling result shown in Figure 6.4. Moreover, the final mudline height with
5 mM Ca2+ addition was much lower than other cases, which is about 0.326.
96
Figure 6.6 Proposed aggregate structures in 10 mM KCl solutions with 0, 0.1 mM Ca2+
addition (a), with 0.5 and 1 mM Ca2+ addition (b), and with 5 mM Ca2+ addition (c).
According to the Stern potential study in Chapter 4, the Stern potentials of the three
kaolinite surfaces were all negative at pH 8 when there were 0, 0.1 mM addition of calcium
ions added into 10 mM KCl solutions, so the kaolinite particles were repelling to each other
in such suspensions as shown in Figure 6.6 (a). When 0.5 and 1 mM Ca2+ addition highly
reduced the negative Stern potentials or even caused charge reversal of Al-basal and edge
surfaces, card-house structure shown in Figure 6.6 (b) would reappear. When more calcium
ions were added, the dominating attraction started to come from Al-basal planes vs edge
surfaces, and from Si-basal planes vs edge surfaces, thus forming a brush-like structure as
we can see from Figure 6.6 (c). Faster settling was observed when the addition of Ca2+ was
increased as predicted, which is because of the compression of edl increased with the
concentration of divalent cations, and the repulsive forces among particles were gradually
reduced. The fast settling achieved at 5 mM Ca2+ addition is caused by the formation of the
compact brush-like structure.
Settling tests were also performed in 5 mM addition of Mg2+ at pH 8 and 11 to see how the
surface precipitation impacts the particle stacking style in a suspension. The results are
shown in Figure 6.7.
97
Figure 6.7 Impact of Mg2+ on the settling performance of kaolinite suspension.
As we can see from Figure 6.7 that the settling performance of kaolinite suspension in 10
mM KCl solution with 5 mM Mg2+ addition is very similar to the settling performance
observed with the same concentration of Ca2+ addition. However, when the pH of
electrolyte solutions was increased to 11, the ISR increased a little for Ca2+ addition due to
stronger EDL compression. As anticipated, due to more formation of magnesium
precipitate at high concentration of Mg2+ dosage at pH 11, all three kinds of kaolinite
surfaces started to have the electrokinetic features of magnesium precipitate. Thus, the
repelling forces among kaolinite particles increased, resulting in a smaller ISR.
98
Chapter 7 Conclusions and Future Work
7.1 Conclusions
In this research, the influence of divalent cations (Ca2+ and Mg2+) on the anisotropic
charging characteristics of kaolinite were thoroughly studied. The new information on the
Stern potentials of kaolinite surfaces in 10 mM KCl solutions in the presence of Ca2+ and
Mg2+ gives insights into their adsorption mechanisms on distinct kaolinite surfaces. Direct
force measurements using AFM on kaolinite basal planes and edge surfaces described in
this research can be extended to other phyllosilicates to explore their surface properties and
colloidal behaviors, which could lead to great contributions to the processing and
utilization of broad natural resources.
To further investigate the ion adsorption mechanisms, atomic resolution imaging was
performed. As a sophisticated technique, atomic resolution imaging is of great application
potentials in colloids and interface field. A detailed operation procedure was described in
this work. In-situ observation of ion distribution at the solid-liquid interface allowed us to
understand the interaction between divalent cations and kaolinite surfaces. When calcium
ions were added into the background solution at high pH, we could see from the in-situ
images that kaolinite surface lattice changed from hexagonal rings to rectangular arrays,
which was caused by the strongly and specifically adsorbed Ca(OH)+.
Finally, the calculation of the interaction energy between kaolinite surfaces explained the
macroscopic behaviors of kaolinite suspension under different solution conditions. The
success link between the microscopic world and the macroscopic world not only provides
99
valuable fundamental guidance for industry application, but also proves that our methods
used in the Stern potential study are feasible and reliable.
In conclusion, the major contributions of this work are:
1) in this study, the method used to obtain the Stern potential of a sample surface based on
ion adsorption mechanism were discussed in detail, which can lead to more accurate Stern
potential results;
2) based on the knowledge of the Stern potential of phyllosilicate clay, the aggregate
structure of the clay particles can be predicted by calculating the interaction energies
among the interacting surfaces. This not only laid a good foundation for future research in
tailings treatment, but can also be used to other research areas, such as controlling nano-
structure of catalyst, synthesizing structural materials, etc.
3) with the presence of Ca2+ in the electrolyte solution of high pH, in situ imaging of the
ion distribution at kaolinite-electrolyte interface directly showed the existence of specific
adsorption for the first time. This technique can also be applied to explore the interaction
mechanisms between reagents and mineral surfaces (e.g., the effect of citric acid on
reducing coagulation of bitumen with kaolinite), to investigate the impact of ambient
conditions on the nanostructure features of material surfaces (such as membranes,
semiconductors, etc.), and so on.
7.2 Recommendations for future work
In this study, the force fitting process can only achieve good fit at the separation
distance > 3 nm from the surface. To obtain a better fit, future research may
100
consider including the repulsive non-DLVO hydration force into the calculation of
theoretical interaction force:
𝐹 = 𝐹𝑣𝑑𝑤 + 𝐹𝑒𝑑𝑙 + 𝐹ℎ𝑦𝑑𝑟𝑎𝑡𝑖𝑜𝑛
The calculation of the hydration force suitable for this AFM tip-solution-substrate
system has been provided by Drelich et al. [75]
Calculate the surface charge density of kaolinite basal surfaces when divalent
cations were added into the 10 mM KCl solution. Use the impact of the divalent
cation on the surface charge density to prove the feasibility of obtaining the surface
lattice structure change seen in Chapter 5.
When 1 mM Ca2+ added into the background solutions of pH 11, in order to confirm
the change of the crystal lattice structure observed on Si-basal plane and Al-basal
plane of kaolinite was caused by specifically adsorbed calcium ions, TOM-SIMS
or DFT calculation could be done in the future to provide further information.
Determine the ratios of solid to cation used in the Stern potential study, and
calculate the usage of cation for settling test using the same solid to cation ratios.
Then conduct settling tests again to see how divalent cations affect the settling
behavior of kaolinite suspension.
Divalent cations can affect kaolinite aggregate structure by influencing the
thickness of electrical double-layer and the Stern potentials of kaolinite surfaces.
To find the link between the divalent cation and the kaolinite aggregate structure,
the settling performance of kaolinite suspension should be investigated under
different concentrations of divalent cations (from 1 mM to 10 mM). This link can
be used to find the optimum dosage of divalent cations for achieving fast settling.
101
Bibliography
(1) WATER MANAGEMENT IN OIL SANDS MINING FACILITIES.
http://www.oilsandsmagazine.com/technical/mining/water-management?rq=water
(accessed in Nov. 19, 2016).
(2) Masliyah, J. H. Bitumen Extraction, Extraction and Upgrading of Oil Sands
Bitumen. Keyano College, Fort McMurray, AB 2003.
(3) Wang, L.; TRONG, D.-V.; Xu, Z.; Masliyah, J. H. Use of Short-Chain Amine in
Processing of Weathered/oxidized Oil Sands Ores. Energy & Fuels 2010, 24
(MAIJUN), 3581–3588.
(4) Liu, J.; Xu, Z.; Masliyah, J. Colloidal Forces between Bitumen Surfaces in
Aqueous Solutions Measured with Atomic Force Microscope. Colloids Surfaces A
Physicochem. Eng. Asp. 2005, 260 (1), 217–228.
(5) Takamura, K.; R.S., C.; D.L., T. The Prediction of Electrophoretic Mobilities and
the Coagulation Behavior of Bitumen-in-Water Emulsions in Aqueous NaCl and
CaCl2 Solutions Using Ionizable Surface-Group Model. In Proceedings of the
Symposium on Flocculation in Biotechnology and Separation Systems; San
Francisco, CA.
(6) Liu, J.; Xu, Z.; Masliyah, J. Studies on Bitumen-Silica Interaction in Aqueous
Solutions by Atomic Force Microscopy. Langmuir 2003, 19 (9), 3911–3920.
(7) Zhao, H.; Dang-Vu, T.; Long, J.; Xu, Z.; Masliyah, J. H. Role of Bicarbonate Ions
in Oil Sands Extraction Systems with a Poor Processing Ore. J. Dispers. Sci.
Technol. 2009, 30 (6), 809–822.
(8) Masliyah, J.; Zhou, Z.; Xu, Z.; Czarnecki, J.; Hamza, H. Understanding Water-
Based Bitumen Extraction from Athabasca Oil Sands. Can. J. Chem. Eng. 2004,
82 (4), 628–654.
(9) Takamura, K.; Wallace, D. The Physical Chemistry of the Hot Water Process. J.
Can. Pet. Technol. 1988, 27 (6).
(10) Takamura, K.; Wallace, D. Experimental and Theoretical Studies of the Hot Water
Processability of Different Grades of Athabasca Oil Sands. In Flocculation in
biotechnology and separation systems; Attia, Y. A., Ed.; Elsevier Science:
Amsterdam, The Netherlands, 1987; pp 579–597.
(11) Hepler, L. G.; Smith, R. G. The Alberta Oil Sands: Industrial Procedures for
Extraction and Some Recent Fundamental Research, AOSTRA Technical
Publication Series# 14. Alberta Oil Sands Technol. Res. Authority, Edmonton, AB
1994.
102
(12) Fong, N.; Ng, S.; Chung, K. H.; Tu, Y.; Li, Z.; Sparks, B. D.; Kotlyar, L. S.
Bitumen Recovery from Model Systems Using a Warm Slurry Extraction Process:
Effects of Oilsands Components and Process Water Chemistry. Fuel 2004, 83 (14–
15), 1865–1880.
(13) Zhao, H.; Long, J.; Masliyah, J. H.; Xu, Z. Effect of Divalent Cations and
Surfactants on Silica-Bitumen Interactions. Ind. Eng. Chem. Res. 2006, 45 (22),
7482–7490.
(14) Basu, S.; Nandakumar, K.; Lawrence, S.; Masliyah, J. Effect of Calcium Ion and
Montmorillonite Clay on Bitumen Displacement by Water on a Glass Surface.
Fuel 2004, 83 (1), 17–22.
(15) Takamura, K.; Chow, R. S. A Mechanism for Initiation of Bitumen Displacement
from Oil Sand. J. Can. Pet. Technol. 1983, 22 (6).
(16) Currie, D. J.; Hathaway, A. P.; Isaacs, E. E.; Mar, A.; Morrison, D. N.; Richmond,
C. The Potential of the Crumble Tests as a Tool for Oil Sands Research. In 30th
Can. Chem. Eng. Conference; 1980; Vol. 4, pp 1125–1142.
(17) Xu, Z.; Masliyah, J. H. Effect of Recycle Water from CT Process on Bitumen
Recovery from Oil Sands, Progress Report #2, A Joint Project with Syncrude,
Suncor Albian and NSERC; 2001.
(18) Kasperski, K. L. Review of Research on Aqueous Extraction of Bitumen from
Mined Oil Sands. Unpublished Report. CANMET Energy Technology Centre,
Natural Resources Canada, Devon, Alta. 2003.
(19) Sobkowicz, J. C. History and Developments in the Treatment of Oil Sands Fine
Tailings. In Keynote address in Proc. 14th International Conference on Tailings
and Mine Waste, Vail, Colorado, USA; 2010; pp 11–30.
(20) Masliyah, J. H.; Czarnecki, J.; Xu, Z. Handbook on Theory and Practice of
Bitumen Recovery from Athabasca Oil Sands - Volume 1: Theoretical Basis; 2011.
(21) John, S. Oil Sands Tailings Technology Deployment Roadmap Project Report –
Volume 2: Component 1 Results
http://www.cosia.ca/uploads/documents/id10/Tailings Roadmap Volume 2 June
2012.pdf.
(22) Technical Guide for Fluid Fine Tailings Management
http://www.cosia.ca/uploads/documents/id7/TechGuideFluidTailingsMgmt_Aug20
12.pdf.
(23) Guggenheim, S.; Martin, R. T. Definition of Clay and Clay Mineral: Joint Report
of the AIPEA Nomenclature and CMS Nomenclature Committees. Clays Clay
Miner. 1995, 43 (2), 255–256.
103
(24) Bergaya, F.; Theng, B. K.; Lagaly, G. Handbook of Clay Science; Elsevier, 2011;
Vol. 1.
(25) Moore, D. M.; Guggenheim, S.; Martin, R. T. Definition of Clay and Clay
Mineral; Joint Report of AIPEA Nomenclature and CMS Nomenclature
Committees; Comment and Reply. Clays Clay Miner. 1996, 44 (5), 710–715.
(26) Van Olphen, H. An Introduction to Clay Colloid Chemistry. Soil Sci. 1964, 97 (4),
290.
(27) Gupta, V. No Title. Surf. Charg. Featur. kaolinite Part. their Interact. 2011.
(28) Yan, L.; Alberta, U. of. Study of Anisotropic Surface Property of Phyllosilicates by
Atomic Force Microscopy; 2013.
(29) Sondi, I.; Bišćan, J.; Pravdić, V. Electrokinetics of Pure Clay Minerals Revisited.
J. Colloid Interface Sci. 1996, 178 (2), 514–522.
(30) Duc, M.; Gaboriaud, F.; Thomas, F. Sensitivity of the Acid–base Properties of
Clays to the Methods of Preparation and Measurement: 1. Literature Review. J.
Colloid Interface Sci. 2005, 289 (1), 139–147.
(31) Yeganeh, M. S.; Dougal, S. M.; Pink, H. S. B. T.-Q. E. & L. S. C. Vibrational
Spectroscopy of Water at Charged Liquid/solid Interfaces; 1999.
(32) Hopkins, A. J.; McFearin, C. L.; Richmond, G. L. Investigations of the Solid–
aqueous Interface with Vibrational Sum-Frequency Spectroscopy. Curr. Opin.
Solid State Mater. Sci. 2005, 9 (1–2), 19–27.
(33) Gerber, C. Atomic Force Microscope G. Binnig “a”and CF Quate ‘b. Phys. Rev.
Lett. 1986.
(34) Butt, H.-J.; Cappella, B.; Kappl, M. Force Measurements with the Atomic Force
Microscope: Technique, Interpretation and Applications. Surf. Sci. Rep. 2005, 59
(1–6), 1–152.
(35) Gupta, V.; Miller, J. D. Surface Force Measurements at the Basal Planes of
Ordered Kaolinite Particles. J. Colloid Interface Sci. 2010, 344 (2), 362–371.
(36) Liu, J.; Sandaklie-Nikolova, L.; Wang, X.; Miller, J. D. Surface Force
Measurements at Kaolinite Edge Surfaces Using Atomic Force Microscopy. J.
Colloid Interface Sci. 2014, 420, 35–40.
(37) Eaton, P.; West, P. Atomic Force Microscopy; Oxford University Press, 2010.
(38) Derjaguin, B. V; Landau, L. Theory of the Stability of Strongly Charged
Lyophobic Sols and of the Adhesion of Strongly Charged Particles in Solutions of
Electrolytes. Acta Physicochim. 1941, 14 (6), 633–662.
104
(39) Verwey, E. J. W.; Overbeek, J. T. G. Theory of Stability of Lyophobic Colloids;
Elsevier: Amsterdam, 1948.
(40) Drelich, J.; Long, J.; Yeung, A. Determining Surface Potential of the Bitumen-
Water Interface at Nanoscale Resolution Using Atomic Force Microscopy. Can. J.
Chem. Eng. 2008, 85 (5), 625–634.
(41) Kosmulski, M. Chemical Properties of Material Surfaces; Marcel Dekker: New
York, 2001; Vol. 102.
(42) Yan, L.; Englert, A. H.; Masliyah, J. H.; Xu, Z. Determination of Anisotropic
Surface Characteristics of Different Phyllosilicates by Direct Force Measurements.
Langmuir 2011, 27 (21), 12996–13007.
(43) Long, J.; Xu, Z.; Masliyah, J. H. Role of Illite-Illite Interactions in Oil Sands
Processing. Colloids Surfaces A Physicochem. Eng. Asp. 2006, 281 (1–3), 202–
214.
(44) Yin, X.; Yan, L.; Liu, J.; Xu, Z.; Miller, J. D. Anisotropic Surface Charging of
Chlorite Surfacesa. Clays Clay Miner. 2013, 61 (2), 152–164.
(45) Fuerstenau, M. C.; Han, K. N. Principles of Mineral Processing; Society for
Mining, Metallurgy, and Exploration: Littleton, Colo., 2003.
(46) Yan, L.; Masliyah, J. H.; Xu, Z. Understanding Suspension Rheology of
Anisotropically-Charged Platy Minerals from Direct Interaction Force
Measurement Using AFM. Curr. Opin. Colloid Interface Sci. 2013, 18 (2), 149–
156.
(47) Yin, X.; Yan, L.; Liu, J.; Xu, Z.; Miller, J. D. Anisotropic Surface Charging of
Chlorite Surfaces. Clays Clay Miner. 2013, 61 (2), 152–164.
(48) Alagha, L.; Wang, S.; Yan, L.; Xu, Z.; Masliyah, J. Probing Adsorption of
Polyacrylamide-Based Polymers on Anisotropic Basal Planes of Kaolinite Using
Quartz Crystal Microbalance. Langmuir 2013, 29 (12), 3989–3998.
(49) Lu, Z.; Liu, Q.; Xu, Z.; Zeng, H. Probing Anisotropic Surface Properties of
Molybdenite by Direct Force Measurements. Langmuir 2015, 31 (42), 11409–
11418.
(50) Hu, Y.; Wei, S.; Hao, J.; Miller, J. D.; Fa, K. The Anomalous Behavior of
Kaolinite Flotation with Dodecyl Amine Collector as Explained from Crystal
Structure Considerations. Int. J. Miner. Process. 2005, 76 (3), 163–172.
(51) Miller, J. D.; Nalaskowski, J.; Abdul, B.; Du, H. Surface Characteristics of
Kaolinite and Other Selected Two Layer Silicate Minerals. Can. J. Chem. Eng.
2008, 85 (5), 617–624.
105
(52) Cuddy, G. Oil Sands Geology. Guest Lecture Notes for Chemical Enginnering
534, Fundamentals of Oil Sands Extraction, January 7-9, at University of Alberta,
Edmonton; 2004.
(53) Liu, J. Surface Properties of Kaolinite Particles - Their Interactions and Flotation
Consideration, The University of Utah, 2015.
(54) Yin, X.; Gupta, V.; Du, H.; Wang, X.; Miller, J. D. Surface Charge and Wetting
Characteristics of Layered Silicate Minerals. Adv. Colloid Interface Sci. 2012,
179–182, 43–50.
(55) Tombácz, E.; Szekeres, M. Surface Charge Heterogeneity of Kaolinite in Aqueous
Suspension in Comparison with Montmorillonite. Appl. Clay Sci. 2006, 34 (1–4),
105–124.
(56) Alkan, M.; Demirbaş, Ö.; Doğan, M. Electrokinetic Properties of Kaolinite in
Mono-and Multivalent Electrolyte Solutions. Microporous Mesoporous Mater.
2005, 83 (1), 51–59.
(57) Gupta, V.; Hampton, M. A.; Stokes, J. R.; Nguyen, A. V; Miller, J. D. Particle
Interactions in Kaolinite Suspensions and Corresponding Aggregate Structures. J.
Colloid Interface Sci. 2011, 359 (1), 95–103.
(58) Kosmulski, M. IEP as a Parameter Characterizing the pH-Dependent Surface
Charging of Materials Other than Metal Oxides. Adv. Colloid Interface Sci. 2012,
171, 77–86.
(59) Helmholtz, H. L. F. von. Studies of Electric Boundary Layers. Wied.Ann 1879, 7,
337–382.
(60) Gouy, M. Sur La Constitution de La Charge Électrique À La Surface D’un
Électrolyte. J.Phys.Theor.Appl. 1910, 9 (1), 457–468.
(61) Chapman, D. L. LI. A Contribution to the Theory of Electrocapillarity. London,
Edinburgh, Dublin Philos. Mag. J. Sci. 1913, 25 (148), 475–481.
(62) Stern, O. Zur Theorie Der Elektrolytischen Doppelschicht. Zeitschrift für
Elektrochemie und Angew. Phys. Chemie 1924, 30 (21‐22), 508–516.
(63) Hunter, R. J. Foundations of Colloid Science, Second.; Oxford University Press:
Oxford ;New York, 2001.
(64) Grahame, D. C. The Electrical Double Layer and the Theory of Electrocapillarity.
Chem. Rev. 1947, 41 (3), 441–501.
(65) Modi, H. J.; Fuerstenau, D. W. Streaming Potential Studies on Corundum in
Aqueous Solutions of Inorganic Electrolytes. J. Phys. Chem. 1957, 61 (5), 640–
643.
106
(66) Ardizzone, S.; Formaro, L.; Lyklema, J. Adsorption from Mixtures Containing
Mono-and Bivalent Cations on Insoluble Oxides and a Revision of the
Interpretation of Points of Zero Charge Obtained by Titration. J. Electroanal.
Chem. Interfacial Electrochem. 1982, 133 (1), 147–156.
(67) James, R. O.; Healy, T. W. Adsorption of Hydrolyzable Metal Ions at the Oxide—
water Interface. I. Co (II) Adsorption on SiO2 and TiO2 as Model Systems. J.
Colloid Interface Sci. 1972, 40 (1), 42–52.
(68) James, R. O.; Healy, T. W. Adsorption of Hydrolyzable Metal Ions at the Oxide—
water Interface. II. Charge Reversal of SiO2 and TiO2 Colloids by Adsorbed Co
(II), La (III), and Th (IV) as Model Systems. J. Colloid Interface Sci. 1972, 40 (1),
53–64.
(69) James, R. O.; Healy, T. W. Adsorption of Hydrolyzable Metal Ions at the Oxide-
Water Interface. III. A Thermodynamic Model of Adsorption. J. Colloid Interface
Sci. 1972, 40 (1), 65–81.
(70) Gan, Y.; Franks, G. V. High Resolution AFM Images of the Single-Crystal α-
Al2O3 (0001) Surface in Water. J. Phys. Chem. B 2005, 109 (25), 12474–12479.
(71) Gan, Y.; Wanless, E. J.; Franks, G. V. Lattice-Resolution Imaging of the Sapphire
(0 0 0 1) Surface in Air by AFM. Surf. Sci. 2007, 601 (4), 1064–1071.
(72) Mugele, F.; Bera, B.; Cavalli, A.; Siretanu, I.; Maestro, A.; Duits, M.; Cohen-
Stuart, M.; Van Den Ende, D.; Stocker, I.; Collins, I. Ion Adsorption-Induced
Wetting Transition in Oil-Water-Mineral Systems. Sci. Rep. 2015, 5, 10519.
(73) Siretanu, I.; Ebeling, D.; Andersson, M. P.; Stipp, S. L. S.; Philipse, A.; Stuart, M.
C.; Van Den Ende, D.; Mugele, F. Direct Observation of Ionic Structure at Solid-
Liquid Interfaces: A Deep Look into the Stern Layer. Sci. Rep. 2014, 4, 4956.
(74) den Ende, D. Atomic Structure and Surface Defects at Mineral-Water Interfaces
Probed by in Situ Atomic Force Microscopy. Nanoscale 2016, 8 (15), 8220–8227.
(75) Drelich, J.; Long, J.; Xu, Z.; Masliyah, J.; White, C. L. Probing Colloidal Forces
between a Si3N4 AFM Tip and Single Nanoparticles of Silica and Alumina. J.
Colloid Interface Sci. 2006, 303 (2), 627–638.
(76) Hamaker, H. C. The London—van Der Waals Attraction between Spherical
Particles. physica 1937, 4 (10), 1058–1072.
(77) Lifshitz, E. M. The Theory of Molecular Attractive Forces between Solids. 1956.
(78) Israelachvili, J. N. Intermolecular and Surface Forces: Third Edition; 2011.
(79) Bergström, L. Hamaker Constants of Inorganic Materials. Adv. Colloid Interface
Sci. 1997, 70, 125–169.
107
(80) Israelachvili, J. N. Intermolecular and Surface Forces, Third.; Academic Press:
Burlington, MA, 2011.
(81) Graf, K.; Kappl, M. Physics and Chemistry of Interfaces; John Wiley & Sons,
2006.
(82) Sokolov, I.; Ong, Q. K.; Shodiev, H.; Chechik, N.; James, D.; Oliver, M. AFM
Study of Forces between Silica, Silicon Nitride and Polyurethane Pads. Journal of
Colloid and Interface Science. 2006, pp 475–481.
(83) Zhao, H.; Bhattacharjee, S.; Chow, R.; Wallace, D.; Masliyah, J. H.; Xu, Z.
Probing Surface Charge Potentials of Clay Basal Planes and Edges by Direct Force
Measurements. Langmuir 2008, 24 (22), 12899–12910.
(84) Larson, I.; Drummond, C. J.; Chan, D. Y. C.; Grieser, F. Direct Force
Measurements between Silica and Alumina. Langmuir 1997, 13 (7), 2109–2112.
(85) Yan, L.; Masliyah, J. H.; Xu, Z. Interaction of Divalent Cations with Basal Planes
and Edge Surfaces of Phyllosilicate Minerals: Muscovite and Talc. J. Colloid
Interface Sci. 2013, 404, 183–191.
(86) Yin, X.; Drelich, J. Surface Charge Microscopy: Novel Technique for Mapping
Charge-Mosaic Surfaces in Electrolyte Solutions. Langmuir 2008, 24 (3), 8013–
8020.
(87) Warszyński, P.; Adamczyk, Z. Calculations of Double-Layer Electrostatic
Interactions for the Sphere/Plane Geometry. J. Colloid Interface Sci. 1997, 187 (2),
283–295.
(88) Wang, M.; Revil, A. Electrochemical Charge of Silica Surfaces at High Ionic
Strength in Narrow Channels. J. Colloid Interface Sci. 2010, 343, 381–386.
(89) Yang, D.; Xie, L.; Bobicki, E.; Xu, Z.; Liu, Q.; Zeng, H. Probing Anisotropic
Surface Properties and Interaction Forces of Chrysotile Rods by Atomic Force
Microscopy and Rheology. Langmuir 2014, 30 (36), 10809–10817.
(90) Gan, W.; Crozier, B.; Liu, Q. Effect of Citric Acid on Inhibiting Hexadecane-
Quartz Coagulation in Aqueous Solutions Containing Ca2+, Mg2+and Fe3+ions.
Int. J. Miner. Process. 2009, 92 (1–2), 84–91.
(91) Kasprzyk-Hordern, B. Chemistry of Alumina, Reactions in Aqueous Solution and
Its Application in Water Treatment. Advances in Colloid and Interface Science.
2004, pp 19–48.
(92) Binner, J.; Zhang, Y. Characterization of Silicon Carbide and Silicon Powders by
XPS and Zeta Potential Measurement. J. Mater. Sci. Lett. 2001, 20 (2), 123–126.
108
(93) Lin, J. X.; Wang, L. Adsorption of Dyes Using Magnesium Hydroxide-Modified
Diatomite. Desalin. Water Treat. 2009, 8 (1–3), 263–271.
(94) Fuerstenau, M. C.; Miller, J. D.; Kuhn, M. C. Chemistry of Flotation.; Soc of
Mining Engineers of AIME, 1985.
(95) Gupta, V.; Hampton, M. A.; Nguyen, A. V; Miller, J. D. Crystal Lattice Imaging
of the Silica and Alumina Faces of Kaolinite Using Atomic Force Microscopy. J.
Colloid Interface Sci. 2010, 352 (1), 75–80.
(96) Masliyah, J. H.; Bhattacharjee, S. Electrokinetic and Colloid Transport
Phenomena; 2005.
(97) Kar, G.; Chander, S.; Mika, T. . The Potential Energy of Interaction between
Dissimilar Electrical Double Layers. J. Colloid Interface Sci. 1973, 44 (2), 347–
355.
(98) Mpofu, P.; Addai-Mensah, J.; Ralston, J. Influence of Hydrolyzable Metal Ions on
the Interfacial Chemistry, Particle Interactions, and Dewatering Behavior of
Kaolinite Dispersions. J. Colloid Interface Sci. 2003, 261 (2), 349–359.
(99) Gan, W.; Crozier, B.; Liu, Q. Effect of Citric Acid on Inhibiting Hexadecane–
quartz Coagulation in Aqueous Solutions Containing Ca2+, Mg2+ and Fe3+ Ions.
International Journal of Mineral Processing. 2009, pp 84–91.
109
Appendices
Appendix A Atomic resolution image processing
Figure A1. Image processing procedure. (a) Original atomic resolution image after being
flattened, (b) Spectrum 2D image, (c) inverse FFT image, and (d) zoomed-in view of (c).
The atomic resolution image of the kaolinite Si-basal plane taken in Milli-Q water was
used to show the image processing procedure. The original figure was 2nd-order flattened
using NanoScope Analysis Version 1.8 (Bruker Corporation) and shown in Figure A1(a).
Lowpass filtering can be used to further improve the image clarity if noise is high.
Spectrum 2D function was then used to transform the image data to frequency domain via
a 2D fast Fourier transform (FFT), as we can see in Figure A1(b). By passing selected
110
frequencies, the image was filtered and certain surface features (in this case, the surface
lattice structure) were isolated, as shown in Figure A1(c). Finally, a zoomed-in view was
provided to see the surface lattice structure image more clearly.
111
Appendix B Error analysis of the obtained Stern potentials
The AFM experimental force curves and the force fitting used to obtain the Stern potentials
in Chapter 4 may bring about certain errors, which mainly come from following reasons:
1) Surface roughness of the substrate/kaolinite surfaces. The surface roughness of the
kaolinite basal surfaces is between 0.3 ~ 1.2 nm, and the surface roughness of edge surfaces
usually ranged from 1.06 ~ 2.3 nm, which is one of the contributions of the discrepancy
between the experimental force curves and the theoretical force curves at short separation
from the surfaces.
2) Non-DLVO forces were not taken into consideration in this study, such as the short-
range hydration force.
3) Experimental force curves obtained from different locations on kaolinite surfaces may
lead to a variation of the derived Stern potential, and the variation is usually between ±2
to ±8 mV. The variation of a few cases reached to about ±10 mV, which usually happened
near the PZC of the surfaces. This is because when near PZC, even slight change of the
solution condition would cause big fluctuation of the Stern potentials of a surface.
4) In the calculation of the theoretical forces, although the surface charging mechanisms
have been seriously explored and used to decide the studied surfaces carry constant surface
potential or constant surface charge, absolute constant surface potential and constant
surface charge are ideal cases. In reality, the interaction force between AFM tip and the
substrate/kaolinite surface would always fall in between the two extreme cases. Therefore,
112
Calculating the theoretical forces based on surface charging mechanisms can only provide
the best approximated forces to the experimental forces.
5) Error may also come from the determination of the spring constant of the cantilever, the
deflection sensitivity, and the geometry of the AFM tip.
113
Appendix C Speciation diagram
Figure A1 shows the speciation diagram for 10-3 mol/L Ca2+ and Mg2+, which was
constructed using data from Fuerstenau et al.94, and Gan et al.99.
6 7 8 9 10 11 12 13 14
1E-8
1E-7
1E-6
1E-5
1E-4
1E-3Ca
2+
Ca(OH)2 (S)
Ca(OH)+
Co
nce
ntr
atio
n (
mo
l/L
)
pH
Mg(OH)+
Mg2+
Mg(OH)2 (S)
Figure C1. Concentration diagram for 10-3 mol/L Ca2+ and Mg2+ .